Wednesday, January 13, 2010


Me, Inflation, and Crazy Pills

[UPDATE below.]

One of my favorite Will Ferrell lines is from Zoolander when his character is exasperated that others can't see that supermodel Dereck Zoolander's "looks" are all the same. He screams, "Doesn't anyone notice this? I feel like I'm taking crazy pills!"

It is with such an attitude that I read this piece from the WSJ (HT2EPJ):
Those looking to Wednesday's "beige book" report for clear signs of a turning point in the Federal Reserve's interest-rate policy might end up disappointed.
High unemployment is keeping a lid on wages, and manufacturers are operating less than 70% of their capacity.

Friday's consumer-price report is expected to show that 2009 marked the first annual decline in consumer prices in 54 years.

The disappointing December jobs report, which showed the loss of another 85,000 nonfarm jobs last month and a 10% unemployment rate, is another reminder that even with vigorous GDP growth, the U.S. economy is operating well below capacity.

"The risks of tightening too early ... still seem bigger than the risks of tightening too late and seeing inflation rise," Goldman Sachs chief economist Jan Hatzius said in a recent client note.

As a result, Mr. Hatzius says he expects the fed-funds rate to stay put near zero this year, "and more likely than not, in 2011 as well." Officials are unlikely to rethink that stance until inflationary gauges start showing persistent gains, economists say.

For now, with inflation running so low, even a much more bullish beige book wouldn't mean rate rises are coming soon.

OK kids, I am going to feel like a serious moron if I turn out to be wrong since I'm harping on it so much, but WHAT THE HECK. I thought the whole lesson of the 1970s stagflation was that the standard Keynesian view--that there is a tradeoff between high unemployment and high price inflation--was totally discredited. That's what I taught my macro students at Hillsdale College, and I thought the WSJ writers were on board with the story.

As far as the speculation that consumer prices in 2009 will register their first drop since 1954, well, go watch the Will Ferrell clip. I don't have to go to ShadowStats to come up with the "real" inflation numbers, I'll go to the St. Louis Fed's website which compiles the BLS' own numbers. Here are some interesting facts:

* Using the normal CPI without seasonal adjustment, prices from Dec 2008 through Nov 2009 (i.e. the first 11 months' of price changes in calendar 2009) increased 2.9%.

* Using the seasonally adjusted CPI, prices in the first 11 months of 2009 went up 2.7%.

* Using the "core" CPI (i.e. taking out food and energy prices) but without seasonal adjustment, "prices" rose 2.0% in the first 11 months of 2009.

* Finally, using the seasonally adjusted core CPI, prices rose 1.7% in the same period.

Note that I am NOT bumping up the percentage increases to annualize the growth rates. No, the numbers above show the simple growth over the 11-month period.

Now you might say, "Oh, well people didn't spend a lot this Christmas, so maybe word on the street is that prices fell up to 2 percentage points from November to December?"

Well, that's technically possible, but even in 2008--amidst the world-is-ending panic--the biggest one-month drop in non-seasonally adjusted CPI was less than 2 percentage points (from October 2008 to November 2008). I'm not checking the other categories but I believe the drops would be lower for them, since (a) a big part of the drop was in energy, meaning the regular CPI would fall harder than the "core" CPI, and (b) prices normally drop in the 4th quarter, so the seasonal adjustment would take some of the actual fall out and smooth it across the year.

So I really don't know what to tell you, kids. If people are expecting the CPI to be down for 2009, come this Friday's announcement, they are going to be as "shocked" as the people forecasting job growth in December.

Someone please explain what I'm missing here. It can't possibly be that I'm the only one to check the numbers through November before shooting my mouth off about what the numbers through December will look like. I even checked to see if maybe they are starting "price increases in calendar 2009" with the CPI number for January, even though that will yield only an 11-month increase. But nope, there's still no way even that number could possibly go negative, unless prices dropped last month more than they did when people thought the world was ending.

UPDATE: Mystery solved. Robert Wenzel emailed the WSJ writer who referred him to an IHS Global Insight report. He called them up, and somebody explained that (barring a big surge in December prices which won't happen) the average CPI in 2009 will be lower than the average CPI in 2008. Hence, one might describe this as saying, "Prices fell in 2009 compared to 2008," an event that hasn't happened in at least 50 years.

OK I'm glad we've at least resolved the mystery, but check out the chart of the CPI. You can see that the reason for this is that prices rose steeply into the summer of 2008, and then crashed in the final quarter. From that valley, they rose fairly steadily throughout 2009.

Well I guess that means the financial press might be able to contain the news that year/year price inflation will be near 3% come Friday's announcement. They can, if they choose, go on and on about how this is an unprecedented collapse in prices. But keep in mind that this statistic is driven entirely by the fall in CPI from October through December 2008. Prices have risen throughout 2009.

FWIW, if you're curious I took the raw CPI numbers and projected them forward, using the (arithmetic) average monthly increase for the first 11 months. By March the 12-month trailing average CPI will be higher than the average CPI in the 12 months of calendar 2008.

I think you're right.

In the case of the typical mainstream economists and financial commentators, they are persuaded that inflation is caused by economic growth (more jobs -> more money), and that the recession causes deflation (less jobs -> less money).

So they see what they want to believe is, rather than what is in fact.
I don't even know why people even have to bring up the 1970s for stagflation--look at virtually every third world country in the world right now. Double digit unemployment, double digit inflation. I can't possibly think of any reason why a serious person would even doubt unemployment-inflation duality.

You will be stunned, as I was, by how the BLS calculates their annual CPI number, but it most likely will be down.
I think you are missing the fact that government economists cannot count. I'm serious. They are doing this in every country. There was four months of what they call "deflation" (falling prices) namely september 2008 - january 2009, after that there was only increasing prices. If you remove the energy component from the US CPI (remember oil falling from waaaaaay up there to waaaaaaay down there?) there was in fact no month at all where prices where going down.
But keep in mind that this statistic is driven entirely by the fall in CPI from October through December 2008.

It's driven by the SPIKE UP before the fall, not the fall. If prices were on average lower in 2008, the CPI would be up in 2009.
Post a Comment

Subscribe to Post Comments [Atom]

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]