Category: Statistics

Not Like Vietnam At All…………

No Comments

After all, we aren’t seeing bogus body counts.

In the Vietnam war, there was a statistical anomaly: numbers ending in 5 and 10 were conspicuously absent, to the tune of millions of times beyond what random chance, IIRC, and it was a lecture almost 30 years ago it was greater than 5σ.

Well, we have another magic number in Afghanistan, 30* (also here and here):

Just how often has the U.S. and NATO killed the Taliban in groups of 30 during 2009? The answer may surprise you:

  • Adnkronos, 12/07/2009: “Up to 30 suspected militants were killed in a NATO airstrike on a Taliban hideout in eastern Afghanistan close to the Pakistani border on Monday. The airstrike targeted the village of Sangar Dara in the mountainous Watapur district of Kunar province , the NATO-led International Security Assistance Force (ISAF) said.”
  • SF Chronicle, 12/04/2009: “Air strikes in two areas of the Mohmand border region killed 30 suspected militants, a military statement said. It said the strikes were “highly successful” but provided no further details, including whether any civilians were hurt.”
  • Xinhua, 11/04/2009: “The military said that the troops have killed 30 more militants during the last 24 hours, bringing the total fatalities to 400, as the operation in the country’s tribal area steadily progressed towards the Taliban strongholds in South Waziristan.”
  • Xinhua, 08/31/2009: “At least 30 bodies of suspected Taliban fighters were recovered in northwest Pakistan’s insurgency-hit Swat valley on Monday, witnesses said. The Pakistani army said they were killed in fighting with the security forces.”
  • Calgary Times, 07/04/2009: “The attack included an attempted suicide truck bombing of the base in the Zirok district of southeastern Paktika province, local officials said. As many as 30 Taliban insurgents might have been killed when troops called in air strikes, they said.”
  • Khaleej Times, 06/24/2009: “Thirty Taliban militants were killed in clashes with NATO and Afghan forces in separate incidents in southern Afghanistan, officials said Wednesday.”
  • Straits Times, 06/15/2009: “Security officials in the region said that about 30 militants were killed in Mohmand agency, close to the provincial capital Peshawar.”
  • Monsters and Critics, 05/28/2009: “In another incident, the Afghan Defence Ministry said Thursday that its troops, backed by international forces, killed 30 suspected militants in neighbouring Khost province Wednesday after the militants attacked their joint base.”
  • Monsters and Critics, 05/14/2009: “At least 30 Taliban fighters were killed Thursday when government artillery fire destroyed their hideout in north-west Pakistan, residents and officials said, as concerns about the fate of thousands of refugees in the region grew amid an escalating humanitarian crisis. Up to 30 suspected militants were in the compound when it was hit, and the Taliban have moved the dead and injured to an undisclosed location, he said.”
  • Reuters, 01 April 2009: “U.S. and Afghan forces have killed 30 Taliban fighters, including a local commander, in an operation in Afghanistan’s southern province of Helmand, the Interior Ministry said on Wednesday.”
  • IRNA, 02/17/2009: “Suspected US drone fired missiles on a training camp of Taliban militants in a Pakistani tribal region on Monday, killing around 30 people, witnesses and official sources said.”
  • New York Times, 01/01/2009: “On Wednesday, the Taliban came for revenge. A group of about 30 Taliban fighters swooped in on Mullah Salam’s house and opened fire. They killed at least 20 of his bodyguards, Afghan officials said. The Taliban claimed that they killed 32. Two of the attackers died.”

It appears that this is a holdover from Rumsfeld days:

We don’t know much about how it works, but in 2007, Marc Garlasco, the Pentagon’s former chief of high-value targeting, offered a glimpse when he told Salon magazine that in 2003, “the magic number was 30.” That meant that if an attack was anticipated to kill more than 30 civilians, it needed the explicit approval of then-Defense Secretary Donald H. Rumsfeld or President George W. Bush. If the expected civilian death toll was less than 30, the strike could be OKd by the legal and military commanders on the ground.

We let me quote a decent man who had a complete sh$# for a son, George Romney, Mitt’s dad, who described the misinformation handed to him by the military on Vietnam as “brainwashing.”

He was right, and Obama has allowed himself to buy into this crap.

*Yeah, there is also that whole #3 person on the chain of command of al Qaeda, who has a life span that rivals that of Ensign Liebowitz of star ship security on a landing party with Kirk, Spock, and McCoy.

The Big Story that is Not a Big Story

No Comments

It turns out that there is a significant inaccuracy in GDP figures, it has to do with the way that imports are accounted for in GDP:

The fundamental shortcoming is in the way imports are accounted for. A carburetor bought for $50 in China as a component of an American-made car, for example, more often than not shows up in the statistics as if it were the American-made version valued at, say, $100. The failure to distinguish adequately between what is made in America and what is made abroad falsely inflates the gross domestic product, which sums up all value added within the country.

American workers lose their jobs when carburetors they once made are imported instead. The federal data notices the decline in employment but fails to revalue the carburetors or even pinpoint that they are foreign-made. Because it seems as if $100 carburetors are being produced but fewer workers are needed to do so, productivity falsely rises — in the national statistics.

“We don’t have the data collection structure to capture what is happening in a real time way, or what is being traded and how it is affecting workers,” said Susan Houseman, a senior economist at the W.E. Upjohn Institute for Employment Research in Kalamazoo, Mich., who has done pioneering research in the field. “We have no idea how to measure the occupations being offshored or what is being inshored.”

In terms of GDP, this is, for now at least, a pretty small part of the picture, well under 1%, which makes it a small story.

On the other hand, one of the arguments for offshoring is that by shipping jobs to China, where worker and environmental protections are weak, and an under valued currency further subsidizes these imports, is that it allows us to focus on what we are good at, and thus boost productivity and GDP.

The bottom line is, as William Alterman, the assistant commissioner for international prices at the BLS notes, “What we are measuring as productivity gains may in fact be changes in trade.”

This is a big part of the story, because the argument for free trade is that it creates, or at least increases, overall well being in our society.

The problem is that the delta from free trade may be grossly overstated, or not exist at all.

A Coda on the Bank Failures

No Comments

If you plot annual rate of bank failures vs time of year, it looks like we might hit 150 bank failures this year.

With 7 Fridays to go, they won’t be doing a closing on Christmas, that’s an 5 failures a week until the end of the year.

My guess would be somewhat less, in the 135-145 failures through the end of the year though.

Note that the data at the start of the year is of a smaller grouping, so inherently noisier.

The trend from June on seems to be clear: that bank failures, which were at around 1½ a week, began accelerating, and it now looks to be averaging around 3/week.

Economics Update

No Comments
Click for full size

Too True!
H/t Calculated Risk,
The Artist should have his website up shortly


Fannie Mae Single Family Delinquencies..OUCH

You know with this recession being over and all, maybe someone should tell the consumer, because consumer spending fell by 0.5% in September, the biggest drop in 9-months.

So consumers are skittish, as a new consumer sentiment survey, this time the Reuters / University of Michigan Survey of Consumer Sentiment Survey, fell in October, down to U Michigan survey, 70.6 from 73.5 in September.

So, that’s like 3 different consumer sentiment surveys that I’ve seen in the past 3 days, one up, and two down.

You have permission to be confused.

There are still a lot of people hurting out there, as shown by the Fannie Mae single family delinquency numbers for August. (see graph pr0n)

I am not seeing even a smidgen of a moderation there.

In the central bank world, the banks appear to be slowly walking back from the extreme measures that they took a year ago, with the Federal reserve re-instituting regulations that it suspended which allowed banks to supply capital to affiliates, which is generally a no-no, and the Bank of Japan is slowly pulling out of the credit markets.

Basically, they are trying to slow-walk their quantitative easing (printing money) measures.

It does not mean that they will be raising rates soon, but it does mean that there is a very gradual tightening of money going on.

In any case, the consumer spending numbers have rattled the markets, pushing US treasuries higher.

In stocks, the VIX, an index of stock volatility spiked upward by 24%, which indicates that market participants are expecting major swings in the stock market.

The bearish news today also pushed oil down, on demand concerns, and pushed the dollar up, on a flight to safety.

Economics Update

No Comments

Remember yesterday, when I said that consumer confidence fell? Well, that was the Conference Board. According to Nielsen, U.S. consumer confidence is up for the first time since 2007, as well as most of the rest of the world.

I think that both organizations conduct reputable surveys, but they got different answers because they asked different questions. This is something that one should consider for any survey.

In the world of slightly more objective metrics, we have durable goods orders rising for the 4th time in 6 months, which is good news, but New home sales unexpectedly fell.

I’m not sure why new home sales falling was “unexpected”. They are recorded when the contract is made, and not when they close, whereas existing home sales are recorded at closing, which means that people who had not bought new homes by the end of August, were really pushing it to qualify for the first time buyer tax credit, which require that the deal be closed by the end of November.

The end of the tax credit is why mortgage applications fell, even though rates fell.

In fact the divergence between new and existing home sales (more later) is a real indicator of how much that tax credit is goosing things.

In the world of central banks, the Norwegian central bank raised its benchmark rate, but the New Zealand bank kept its rate steady.

Of course, there is some apples and oranges here, because Norway raised its rate to 1.5%, and the Kiwis kept their rate steady at 2.5%.

In either case, the markets are not being optimistic, with oil falling below $78/bbl, and the dollar and yen strengthening on a flight to safety.

Economics Update (a Day Late)

No Comments
Click for full size

Unemployment Numbers, Actual v. Seasonably Adjusted


Philly Fed Graph Pr0n Courtesy Calculated Risk


NY Fed Graph Pr0n Courtesy The Bonddad Blog

So, Seasonally adjusted first time unemployment claims fell to 514,000, the lowest level since January, the 4 week moving average fell by 9K to 531,500, and continuing claims fell 75K to 5.99 million, the first time that the number has been below 6 million in 6 months.

Well, sort of anyway. As Brad Delong notes, the non-seasonally adjusted number actually went up:

Unemployment Insurance claims rose from 452,000 last week to 504,000 this week, but the seasonal adjustment factor fell from +72,000 to +10,000, leaving seasonally-adjusted claims falling from 524,000 to 514,000.

Considering the strangeness of the times that we are currently going through, this does mean that the SA numbers have a bit of flakiness.

Still these numbers, as well as the New York and Philadelphia Federal Reserve activity indices are definitely trending better.

The reason that I think that this is a pause, rather than a recovery, is because the underlying problems remain unresolved, with foreclosures hitting an all time high in the 3rd quarter.

About 1 out of 136 homes got a foreclosure notice in the past quarter.

That along with the fact that the CPI numbers are showing that “Owners’ Equivalent Rent” is falling, which implies that home prices have even farther to fall before the rent/own ratio is back to where it should be imply to me that the real estate crash is still on the down slope.

Additionally, it’s clear that consumers are still stretched, with Capital One credit card defaults rising in September.

30 year fixed mortgage rates remain below 5%, though they are up a bit this week.

In energy, oil is now at a 2009 high, and in currency, the
dollar rose against the Yen, but fell against the Pound Sterling and Euro.

Birth-Death Adjustment Finally Coming Under Scrutiny

No Comments
Click for full size


Houston, we have a problem.


And job recoveries are progressively slower too

This is kind of a wonky bit about employment statistics in the US, and how a statistical tool, the birth death adjustment, may not be a reasonable way of looking at employment in the United states, and now New York Times columnist Floyd Norris is taking note of the fact that job losses in 2008 are now understood to be far greater than originally reported at the time:

It now appears that during the first half of 2008, when the recession was getting under way, job losses averaged 146,000 per month. That is nearly three times the average of 49,000 jobs shown in the initial estimates.

How did the government get it so wrong?

(emphasis mine)

The answer is very simple, a statistical correction called the “birth-death adjustment”, which is about birth and death of new businesses, rather than the birth and death of people, and it, “factors in jobs assumed to have been created by employers who are too new to have been included in the survey, and subtracts jobs from employers assumed to have failed and therefore not responded to the latest survey”.

You see, under George W. Bush and His Evil Minions™, the birth-death adjustment was massively expanded, just in time to create for the 2004 election. So there are a number of reasons for this:

  • It created better job numbers, and hence political advantage for Bush and His Evil Minions™.
  • It was part of the ideology of the “ownership society” that there were millions of people chomping at the bit to become entrepreneurs, which leads to a feeling that the Birth/Death numbers need to be expanded.
  • Political advantage.
  • A lack of understanding that Americans have become much less entrepreneurial even in comparison to members of other developed economies, because people are unwilling to rely on privately purchased insurance for their healthcare.
  • Political Advantage.
  • It is bad to present Dear Leader with bad news.

As the top graph shows, something is truly whack here.

Unfortunately, I don’t see this being fixed in the foreseeable future, becausethere is no advantage for Obama/Geithner/Summers to start using more accurate, and hence scarier, numbers.

H/t Barry Ritholtz.

U.S. Unemployment Now as High as Europe

No Comments

Actually the US unemployment rate is higher relative to Europe than the Center for Economic and Policy Research (CEPR) report suggests.

The Europeans have a far more relaxed count for the unemployed, counting, for example, full time students looking for part time work as unemployed, and the US counts active duty military as part of the civilian workforce,, which Europe does not, further lowering the US numbers relative to Europe.

When comparing apples to apples, the EU and US have been much closer in terms of unemployment rate, and right now, our unemployment rate is higher.

Additionally, the consequences of unemployment are far worse in the US with a threadbare social safety net which involves, among other things, things like a loss of healthcare.

Republicans Hate Honest Math: Census Edition

No Comments

It appears that Obama selection of Robert Groves as director of the US Census has gotten Republicans worried:

Republicans expressed alarm because of one of Mr. Groves’s specialties, statistical sampling — roughly speaking, the process of extrapolating from the numbers of people actually counted to arrive at estimates of those uncounted and, presumably, arriving at a realistic total.

If minorities, immigrants, the poor and the homeless are those most likely to be missed in an actual head count, and if political stereotypes hold true, then statistical sampling would presumably benefit the Democrats.

First, let me note here that the reporter does not have a clue: As a matter of settled law, the Supreme Court ruled that you cannot use a statistical sampling, but instead must use an actual count in Department of Commerce v. U.S. House of Representatives in 1999, so this is a non-issue.

What is not a non issue is that statistical sampling can find people and places that have been under-sampled, and can so create a more accurate census, and a more accurate census is probably a bad thing™ for Republicans, because white males tend to be the easy folks to count.

The ‘Phants don’t oppose him because he will, “Politicize the Census,” they oppose him because he will do the best job possible within his power.

They don’t want someone to do a good job, they want someone to do a piss poor job.

Economics Update

No Comments

Since we’ve already covered the Geithner political suicide pact economic plan, let’s lead off with housing.

Happy, happy, joy, joy, home sales rose 5.1% in February relative to January, seasonally adjusted, though people not so closely attached to the realty industry have noted that it’s
really a pretty modest rise, or note that home sales have fallen year over year, and that about 45% of these sales are distressed in some manner.

I would say that year over year is the only metric to apply, because the so called “seasonal adjustment” if it ever were valid, has become meaningless in the current collapsing market.

Internationally, we have another member of the 0% benchmark club, as the Bank of Israel has cut its benchmark rate to 0.50%.

In currency and commodities, the dollar is down, because people are less concerned about safe havens, so there is less dollar flight to safety, and copper and oil rose.

It appears that these markets like Geithner’s plans, because, they are over paid and over bonused trader types, I guess.

The Mess That Greenspan Made: The Great Inflation Moderation that Wasn’t

No Comments

Tim at The Mess That Greenspan Made has a great post on how the Federal Reserve, and Alan “Bubbles” Greenspan, succeeded in making a laughing stock of inflation results through the use of imputed rent (owner’s equivalent rent) in the CPI data.

Not only is inflation higher, but the swings in inflation, driven as they are by the aggressive policies of the Federal Reserve to drive down wages manage inflation, are much more extreme.

In essence, as I’ve noted before the Fed has made itself look good.

BTW, you can get more information, in mind numbing detail, from Shadow Stats.

Another Wingnut New Deal Myth Busted

2 Comments

I am not talking here about the delusional ramblings of people like wingnut welfare recipient* Amity Shlaes, who suggests that capitalists put their money in their mattresses in Atlas Shrugged style in 1937 because of the “uncertainties” that FDR produced.

After all, that one has been debunked by a legion of economist including right wing darling Milton Friedman, with the only dispute being to what degree the cause was fiscal tightening in 1937 by FDR (Keynesian) or monetary tightening by the Fed (Chicago School).

Instead, I am talking about the suggestions by the Chicago School boys that the New Deal did little or nothing to fix unemployment.

As economist James Galbraith so ably notes, the people who want to show this do so by not counting employed people.

Actually, it’s more than that, it’s by counting the employed as unemployed.

He starts out with the figures must usually used by neoliberal economists, in which Stanley Lebergott seems to show that the 1937 recession returned unemployment to nearly its prior level.

It seems to be a damning indictment of the new deal, except, as Michael R. Darby noted in 1976, 3½ million employed individuals are not being counted in this graph.

As Lebergott notes in his own footnotes, the people on “Emergency Employment,” things like the WPA and NRA (not the gun lobby) projects are simply not counted, because that is how the BLS counted employment in the 1930s.

What’s more, not only are they not counted, they are counted as unemployed, which paints a very different picture, with the dashed line showing the unemployment numbers.

The critics of FDR are figuring unemployment numbers by using employment data ex-employment data.

While there may be a case, though I would disagree, that these people should be dropped from the employment rolls for counting purpose, as active duty military were until 1982, when they were added to artificially keep the unemployment rate below 10% by Ronald Reagan’s BLS.

It turns out that if you take their argument, which is that government employment isn’t real, and that private sector is what matters, you get a picture which an even more drastic drop in unemployment during the New Deal, in part because private sector unemployment did not peak at 25%, but because it peaked at 30%. (!)

So, now that we have dealt with that shibboleth, I actually found another interesting bit of information, from a handy chart of GDP and federal spending from 1930 to 1984, which allows me to do this GDP chart in both absolute and real (inflation adjusted) terms using CPI data available online.

It shows a fairly blistering decent growth rate in 1934, with blistering double digit growth in 1935 and 1936, and in 1937, the economy, in real terms, finally passes the 1930 economy, as opposed to 1941 in absolute terms.

This gives us an insight as to why Roosevelt listened to his Treasury Secretary, and cut spending: he thought, or at least his gut was telling him, that the economy had grown beyond the level of 1930.

*She graduated with a degree in English from Yale’s Jonathan Edwards College†, never, as near as I can determine, has published anything on economics in a peer reviewed setting, and yet she is a senior fellow in economic history at the Council on Foreign Relations.‡
†Ivy League Welfare too, I guess.
‡The fact that her rise to prominence followed her 1988 marriage to right wing luminary “journalist” Seth Lipsky, former editor-in-chief and president of the faux “newspaper” the New York Sun seems to indicate that nepotism is involved, but compared to Bill Kristol, who would be flipping burgers but for who his father was, it’s a minor example.§
§Why doesn’t the left support its ideological foot soldiers in the same way?

A Better Employment Metric

No Comments

Of course, it is also a very scary metric.

But I agree with Paul Krugman, workforce participation is one of the better metrics that we have to describe the unemployment situation.

However, I do have a problem with the graph, which is that the delta is magnified, by focusing on a range, in this case 61%-65%, as opposed to showing the full range from 1-100%.

I think that these sort of “broken range” graphs serve to overemphasize deltas….though that is a really scary picture, even if it’s done on a full 1-100 scale.

It shows that employment never recovered from the 2001 downturn, and that it’s headed down fast now.

Economics Update

No Comments

Well, I will start of with Mr. Ritholtz’s trenchant analysis of the scary statistics I’ve been throwing about recently. He looks at them and, concludes that the employment numbers are tweaked to make things look less dire.

Basically, he is saying that the birth/death (of small firms, not people) adjustments and the seasonal adjustments are bunk, and the numbers on workforce participation as a percentage of the economy bear him out.

Additionally, you have people leaving the labor force, or being underemployed, which is not counted in the normally covered statistics.

BTW, the fact that foreclosures rose 76% year over year in the 3rd quarter bodes ill for the housing to recover any time soon. (chart pr0n captured by Barry Ritholtz, I highly recommend his site)

Of course, the results of Manpower, Inc. survey of employers does not point to an increasing in hiring in any segment of the economy.

About the only bright news is that looks poised to approve the various bank rescue plans in the Euro zone, they just approved France’s plan.

Going further east, however, things get grim, with S&P downgrading Russian government debt and downgrading 6 of the larger Russian banks.

Despite the news, oil was up today, largely on reports of a bailout of the Big 3 (Big 2½), but retail gasoline is continuing to head down.

Gasoline will continue to lag oil on the way down for 2 reasons: It takes time for oil to get shipped from a refinery and become gasoline, and people are extracting profits on the way down, because what was expensive seems cheap to the consumer now.

That Economic Data is Worse than It Looks

No Comments

One of the common things that I’ve noticed since I’ve started following government economic data is that the preliminary data comes in better than the final data.

When you compare month to month, you are comparing preliminary data to the prior month’s final data, so the delta, which is what the press covers, looks much better than what it is in reality.

Well, Dean Baker just caught a doozy on October industrial production data.

In bullet points:

  • September hurricanes artificially depressed that month’s stats.
  • September data was then revised down.
  • October preliminary data, which showed a “rise” is flat when comparing preliminary to preliminary.
  • If you look at just manufacturing data, which strips out the noisier utility and mining segments, that number is down from September manufacturing data.

What This Means

No Comments

This graph, from The Big Picture, means some combination of the following items:

  1. The economy of the past 18 years has sucked so badly that people have increasingly given up looking up.

  2. The basic unemployment number has been screwed with by administrations on both sides of the aisle, and significantly understates the unemployment rate.

  3. That the economy has shifted significantly in the past 2 decades, and high levels of long term unemployment are the norm.

My money is mostly on number 2, though reverse Robin Hood is part of it.

Short term solution is more aid to the long term unemployed. The long term solution is fixing the BLS data, and creating a more just society.

Truth be told, I’m surprised that the divergence did not occur around 1983, when Reagan and His Evil Minions™ screwed with the unemployment numbers with things like counting active duty military as a part of the workforce, to keep the unemployment number below 10.

Check out Daniel Gross in Slate, who argues that the normal unemployment numbers are complete crap.

Economics Update

No Comments

the producer price index rose 1.2% in July, that comes to about 15% inflation, and the year over year rate was 9.8%.

Inflation is back….Truth be told, it was never gone, it’s just that the government statistics concealed it, and we are now running into the limits of such accounting artistry.

We also are seeing housing starts at a 17 year low, so it looks like stagflation to me.

I just hope that it isn’t an Argentina/USSR style collapse.

I would note that a lot of this inflation is commodities, and they are down.

Both oil and gasoline (33rd straight day) fell again.

That being said, the dollar was down again today. Those inflation numbers probably scared traders.

Finally it looks like Lehman may be forced to sell its money management division in order to raise capital to offset its losses.