Reuters strangely cheered by statistical noise in economic indicators

posted at 10:45 am on October 13, 2011 by Ed Morrissey

The Departments of Commerce and Labor released scheduled reports on the monthly trade deficit (for August) and initial weekly jobless claims today, but I didn’t plan on writing much about either.  They showed no real change from the previous reports in the series, and give no indication one way or the other about the direction of the economy.  That didn’t stop Reuters from offering a relatively cheery analysis, of course, headlined “Claims and trade data suggest modest improvement“.

Let’s start with the actual reports, first on the jobless claims:

In the week ending October 8, the advance figure for seasonally adjusted initial claims was 404,000, a decrease of 1,000 from the previous week’s revised figure of 405,000. The 4-week moving average was 408,000, a decrease of 7,000 from the previous week’s revised average of 415,000.

Last week’s announced figure was 401,000, so this actually represents an increase from that mark — but only a slight one, in the realm of statistical noise.  The decrease from the adjusted figure is even less notable, a change of 0.25%.  It’s essentially a no-change report.  The same is even more true for the trade deficit:

The U.S. Census Bureau and the U.S. Bureau of Economic Analysis, through the Department of Commerce, announced today that total  August  exports of  $177.6 billion and imports of $223.2 billion resulted in a goods and services deficit of  $45.6 billion,  virtually unchanged from July, revised.   August exports were $0.1 billion  less than  July exports of $177.7 billion.   August imports  were $0.1 billion less than July imports of $223.3 billion.

In  August, the goods deficit  increased $0.1 billion from July to $61.4 billion, and the services surplus increased $0.2 billion to $15.8 billion.  Exports of goods  decreased $0.1 billion to $126.7 billion, and imports of goods were virtually unchanged at $188.1 billion.  Exports of services  were virtually unchanged at $50.9 billion,  and imports of services decreased $0.2 billion to $35.1 billion.

The actual change was around $20 million, from $45.63 billion in July to $45.61 billion in August, which represents a change of 0.044% to the trade deficit.  That’s practically a definition of statistical noise.  Exports actually fell slightly, offset by a slightly larger drop in imports, thanks to a stagnant economy.

Now let’s look at how Reuters reported these economic indicators:

New claims for jobless benefits were little changed last week and the trade deficit narrowed marginally in August, indicating a modest improvement in the economy.

Initial claims for state unemployment benefits dipped 1,000 to a seasonally adjusted 404,000, the Labor Department said on Thursday. Economists had expected claims to rise to 405,000.

A separate report from the Commerce Department showed the trade deficit fell to $45.61 billion in August from $45.63 billion the prior month. Economists had expected the trade gap to widen to $45.8 billion.

I’d really like an explanation of how decreases in both exports and imports indicate “a modest improvement in the economy.”  The scale of the drop is so small that no one can really argue that it portends a move to recession, but it’s certainly not a “modest improvement.”

These reports indicate that the national economy continues on its trajectory for stagnation and moribund job-creating activities.  It’s essentially no change at all, and to the extent that there is any change, it’s not encouraging for future growth.  Reuters would have been better off leaving these reports alone.

Blowback

Note from Hot Air management: This section is for comments from Hot Air's community of registered readers. Please don't assume that Hot Air management agrees with or otherwise endorses any particular comment just because we let it stand. A reminder: Anyone who fails to comply with our terms of use may lose their posting privilege.

Trackbacks/Pings

Trackback URL

Comments

The AP and Reuters are charlatans, in Obama’s azz.

Suffocate from what you eat, what you dwell in, derelict fools.

Schadenfreude on October 13, 2011 at 10:48 AM

It’s essentially no change at all, and to the extent that there is any change, it’s not encouraging for future growth.

Ed. You’re absolutely right. By your own historical data, there needs to be at least a change +/- of 9K in order to represent something of significance. Even then, that’s only 2.3%. These folks really do not understand descriptive statistics nor significant change if changes of this “magnitude” actually compel them to sit down and type something about it.

ted c on October 13, 2011 at 10:54 AM

In the Obama world, failure is success and success is pure, god-like, Obama genius.
Anything else is blamed on someone or something else.

albill on October 13, 2011 at 10:59 AM

I never see Captain Pictard’s face these days.

Oil Can on October 13, 2011 at 11:01 AM

Anything to prop up dear leader

cmsinaz on October 13, 2011 at 11:04 AM

Alternative headline:

Rounding error signals economic recovery!

John Deaux on October 13, 2011 at 11:05 AM

Last week’s announced figure was 401,000, so this actually represents an increase from that mark

This seems to happen every week.

Paging Issa?

faraway on October 13, 2011 at 11:06 AM

Reuters is Charlie Brown — albeit it a purposely clueless Charlie Brown — and the first-of-the month national unemployment statistics are Lucy and the football. Reuters “unexpectedly” has to report on lower-than-hoped for job figures the first Friday of every month, then spends the next 30-31 days trying every way in the world to convince itself that next month really is going to be different, and they’re going to finally boot that ball the next time around.

jon1979 on October 13, 2011 at 11:07 AM

up next. Billboard Top 40 includes new fave hit “White Noise” by “The Record Needles”….moving up the chart a blistering pace and being played on record players from coast-to-coast.

I know, I know…dating myself.

ted c on October 13, 2011 at 11:10 AM

Marketing can do many things, but it cannot sell a product that people have tried and rejected. That is Obama’s reelection problem. At the risk of being unsophisticated and abusing the concept of Occam’s Razor, Obama’s reelection problem can be expressed in one simple sentence: “Now, too many people know him.”

Obama’s only strength was Axelrod’s ability to play on the imagination of voters. That strength no longer exists. People now know the product and have rejected it. They did not get even Chauncey Gardiner. Embarrassed and angry, the public is stuck with Chance the Gardener.

The irony is that Mr. Obama has not changed. He is the same man who was elected. His problem is not communicating, Republicans, George Bush, tsunamis, or anything else. His problem is the man in the mirror. There is no more there than an image.

Obama was all hype and no substance. That realization has dawned on voters, resulting in horrendous polling. Richard Nixon was never liked, but he was at least thought competent. Obama was liked but never competent. Now Obama is living proof of the old adage that familiarity breeds contempt. He is neither liked nor competent.

Even the hapless Jimmy Carter did not attain that status.

Schadenfreude on October 13, 2011 at 11:10 AM

but now everything is calculated for ten years (savings of….)

So its really a drop of (405000-404000)x52x10= 520,000!!!!!!

tomg51 on October 13, 2011 at 11:11 AM

I’d really like an explanation of how decreases in both exports and imports indicate “a modest improvement in the economy.” The scale of the drop is so small that no one can really argue that it portends a move to recession, but it’s certainly not a “modest improvement.”

..the same way changing your weekly chocolate ration from 30 grams to 25 grams is an increase. You keep making impudent statements like that, Ed, and you’ll be hauled off to room 101 with a pound of Limburger spread all over your cheeks..

..and, yes, we’re still with Eastasia.

The War Planner on October 13, 2011 at 11:13 AM

..and, yes, we’re still at war with Eastasia.

Drat!

The War Planner on October 13, 2011 at 11:14 AM

Schadenfreude on October 13, 2011 at 11:10 AM

..good column this: The Obama Problem in The American Thinker. Read the whole thing; it’s a tour de force.

The War Planner on October 13, 2011 at 11:17 AM

Ed: whenever data agrees with your hope of
negative economic news, it’s both highly significant and “expected.” But whenever there is good news, it’s “statistical noise.” Methinks you need to apply the same standard of proof to both.

Seven Seas on October 13, 2011 at 11:18 AM

Perhaps, Reuters wants the same thing the commies in D.C. want. The destruction of our economy and country, to be rebuilt in the commie image.

capejasmine on October 13, 2011 at 11:21 AM

Seven Seas on October 13, 2011 at 11:18 AM

I live in the real world. Where do you live?

capejasmine on October 13, 2011 at 11:22 AM

ted c on October 13, 2011 at 10:54 AM

Their analysts have a job to do. They have to write somethin… Come on. what do you want them to do? Join the unemployment lines?

antisocial on October 13, 2011 at 11:27 AM

Ed: whenever data agrees with your hope of
negative economic news, it’s both highly significant and “expected.” But whenever there is good news, it’s “statistical noise.” Methinks you need to apply the same standard of proof to both.

Seven Seas on October 13, 2011 at 11:18 AM

dude, it is totally statistical noise. Ed has this data stored up for every 2 weeks over the last umpteen months. I looked at it myself. For there to be significant, meaningful change in these jobless figures–there must be a change of 9000-9250 claims. A change of 1000 is what we call in statistics a “nothing burger.” Ie, that’s change that is within the realm of chance. If they were to count the claims again, they may come up 1000, or even 2000, or drop by 2500. Either way, it doesn’t matter statistcially until it reaches a certain threshold–that is that there is <5% probability that the change that occurred is due to chance. It hasn't met that threshold.

ted c on October 13, 2011 at 11:31 AM

Either way, it doesn’t matter statistically until it reaches a certain threshold–that is that there is <5% probability that the change that occurred is due to chance. It hasn't met that threshold.

ted c on October 13, 2011 at 11:31 AM

..correct. It was said that the economy needs to produce around 275,000 jobs per month for the next 13 months just to get the unemployment down to 8% and the first-time jobless claim numbers needs to drop to around 375,000 to be anywhere near “normal”.

Another pathetic stat for Mr Seven Seas to chew on: the first-time jobless claim numbers have been above 400,000 for 160 out of the past 168 weeks — 144 of those weeks being The Pantload’s Reign of Error.

Now that’s a significant trend, old son.

The War Planner on October 13, 2011 at 11:54 AM

But whenever there is good news, it’s “statistical noise.” Methinks you need to apply the same standard of proof to both.

Seven Seas on October 13, 2011 at 11:18 AM

-

The decrease from the adjusted figure is even less notable, a change of 0.25%.

-
I have to agree with Seas here Ed… 0.25% is huge…! The OWS crowd can pack up and go back to work now.
-

RalphyBoy on October 13, 2011 at 12:33 PM