Aus dem wöchentlichen Newsletter – Thoughts from the Frontline
von John Mauldin
“There are three kinds of lies: lies, damn lies, and statistics.” – commonly attributed to Benjamin Disraeli
If we are to believe the government statistics, the GDP of the US grew by 0.6% in the first quarter of this year. And unemployment actually fell. And there were only 20,000 job losses. This week we do a quick review of why the statistics can be so misleading. We also look at why I was wrong about the housing number last week, and I highlight what could be a very serious Black Swan lurking in the agricultural bushes. It should make for an interesting letter. It’s hard to know where to begin, there are just so many tempting targets; so let’s take the statistical aberrations in the order they came out this week.
Who Is Inflating the Numbers?
In my January 2007 annual forecast, I said that we would see a recession or a serious slowdown by the end of 2007 and that it would be mild as these things go, triggered by a bursting of the housing bubble and a slowdown in consumer spending. During the summer and specifically in October I wrote that we were facing a Slow Motion Recession – that the recovery process would be lengthy and take several years before we got back to the 3% growth rate that is more typical of the US economy.
There were lots of people who made fun of my forecasts, and some were quite snide. I really let stuff like that roll off my back. But I have yet to see those writers admit they were wrong (as I will at the end of this letter). And I doubt I will. I take little pleasure in being right on the recession call, as recessions are not fun for those in harm’s way, but I call it as I see it. We’ll just have to wait and see if some of my other forecasts come to pass. I am sure I will miss a few things. Part of the nature of the business.
Last week I suggested that this week’s release of the GDP would be slightly positive, as the BEA would have a much lower number for inflation than our common experience suggests to be the case in the real world. It turns out my cynicism was well justified.
The Bureau of Economic Analysis (BEA) of the Department of Commerce publishes the GDP statistics. They tell us the US economy grew by 0.6% in each of the last two quarters. They come by that number by taking the nominal or “current dollar” measure of the economy and subtracting their figure for inflation, which gives us “real GDP,” or after-inflation GDP.
Nominal GDP in the fourth quarter grew by 3%. In the first quarter it was 3.2%. They figure that inflation was 2.4% in the fourth quarter and 2.6% this quarter, giving us the slightly positive growth numbers.
There are several government agencies which track inflation. And in fairness, inflation in an economy as large as that of the US is a very tricky thing to measure. The Consumer Price Index (CPI) is done by another division of the Department of Commerce, the Bureau of Labor Statistics. Let’s look at what they calculate inflation to be since last August, in the following table.
Note the string of five consecutive months of 4%-plus inflation, and that the average for the 4th quarter was 4%, while for the first quarter of 2008 it was over 4.1%. Never mind whether that is the right number or whether there are problems with how they calculate it – that is a story for another letter. The key here is that if the BEA used the BLS number (remember, both groups are in the same Department of Commerce), it would show the economy shrinking by 1% in the 4th quarter and by almost 1% in the first quarter. That is not what the happy-talk analysts are saying.
But let’s use the Fed’s favorite measure of inflation, personal consumption expenditures, or PCE. The PCE has been about one-third less than the CPI since about 1992. The difference is in the way they are calculated. The CPI uses a weighted average of expenditures over several years. As I understand it, the PCE tracks changes in relative expenditures from one quarter to the next, assuming that consumers change their habits as prices rise and fall. In simplistic terms, if steak gets expensive, we substitute with hamburger or chicken. One index tracks those changes over years and the other (PCE) does it over quarters. Also, the PCE only tracks personal consumption and not imports or inventories.
If we use the PCE numbers (yet another measure using Commerce Department data), inflation was about 3.3% for both quarters, which would mean negative growth quarters by a few tenths of a percent. That would also mean two quarters of negative growth and a recession.
Further, GDP in the first quarter was helped by inventory build-up to the tune of 0.8%. In times of expansion it is good to see inventories grow, as that means companies are optimistic. But when the economy begins to slow, growing inventories mean that companies anticipated sales that did not materialize. That means that as inventories are allowed to fall in the second quarter, they will show up as a negative factor in second-quarter GDP.
But all these numbers will be changed in a few years, as looking back over several years is the only way we can get somewhat accurate numbers. My bet is that the numbers for GDP will be revised down when the economy is well on its way to recovery. It will show up on page 16 of the Wall Street Journal and no one will care. That is what happened when we found out a few years later that the last recession started in the third quarter of 2000. The initial numbers were positive.
The “official” arbiter of whether or not we are in a recession is the National Bureau of Economic Research. And they do not use the GDP numbers. If they did, then what would be the point of asking them? We could just look at the government statistics. But we don’t. Normally, we think of two consecutive quarters of negative GDP as a recession. But NBER has other ways to look at it.
Barry Ritholtz sent this note to me:
“The 2 consecutive quarters of GDP contraction is not the only metric for identifying recessions. According to the econo-geeks at the National Bureau of Economic Research, a recession is defined as a “significant decline in economic activity spread across the economy, lasting more than a few months.” Here’s their specific language:
“ ‘Most of the recessions identified by our procedures do consist of two or more quarters of declining real GDP, but not all of them. Our procedure differs from the two-quarter rule in a number of ways. First, we consider the depth as well as the duration of the decline in economic activity. Recall that our definition includes the phrase, ‘a significant decline in economic activity.’ Second, we use a broader array of indicators than just real GDP. One reason for this is that the GDP data are subject to considerable revision. Third, we use monthly indicators to arrive at a monthly chronology.’”
“Hence, if we follow what the people who actually determine what is and isn’t a recession say about the matter, and not just limit our analysis to GDP, then it’s pretty clear we are now experiencing an economic contraction.”
Real (inflation-adjusted) retail sales have been flat for the last six months. Incomes are stagnant. Consumer spending is showing every sign of slowing even more. Unemployment is rising (see more below). Consumer sentiment is at 25-year lows. You can count on it that the NBER will show a recession starting the fourth quarter of last year and continuing at the least through the first quarter of this year. This one could last another six months. I still think long and shallow with a very slow recovery.
One last point. The US population grows by about 1% a year. Thus economic growth should increase by at least 1% for the US to stay even on a per capita basis. Thus, at least with regard to GDP per capita, the US is definitely in a recession. And if you use real-world inflation data, we are also in a mild recession.
Honey, I Blew up the Employment Numbers
Long-time readers know the problems I have demonstrated with the monthly employment report. It is one of the most revised reports released by any government agency, and for some reason the market seems to react to it like it means something immediate.
Let’s take today’s release. It showed a drop of only 20,000 jobs, well above the more negative consensus. The market immediately rallied, taking the thought that the economy may be on its way to recovery. But when you look at the numbers, that optimism evaporates.
The birth/death ratio is the BLS’s attempt to figure out how many jobs were created by small businesses that do not show up in their survey of established businesses. It is a simple estimate based on past trends. You have to have this estimate to have any hope of getting the actual number right. And most of the time, the estimates are pretty good. Over time the numbers are revised and in a few years will be pretty close. But in times when the economy is slowing down, the birth/death ratio tends to overstate job growth because the trend is backward-looking. This month’s birth/death number was particularly egregious.
April, for whatever statistical reason, has shown the highest number of birth/death jobs for any month. In 2007, the BLS estimated that 262,000 were created in April that they could not account for in the survey of businesses. Somehow, the spreadsheets at BLS had them add 267,000 jobs in April of 2008. That number includes an estimated 45,000 new jobs in construction! And this in a time when both residential and commercial construction are contracting. The actual survey results showed that construction jobs fell by 61,000.
And somewhere, they estimate that 8,000 new jobs in finance were created. As Philippa Dunne notes: “It may be that the gains in our old friend, bars and restaurants, are the [birth/death] model’s creation; it added 83,000 to the leisure and hospitality sector. With vacation plans at near-record lows, and restaurants reporting reduced traffic, many of these job gains could disappear in the next benchmark revision.”
Without that addition from the birth/death number, total private employment would have dropped by 296,000. Now, if that had been the headline number, the market would have tanked. Now, I have no doubt that the economy did create a lot of new jobs last month. But when the final revisions are in, we will see that job losses were well south of 100,000. If memory serves me correctly, the BLS had to add about 800,000 jobs that they missed during the recovery in 2003-4. (The birth/death model misses job growth during recoveries, the opposite result of the miss in slowing periods.) They did this just last year, in a major revision of the data. We will see the same type of revisions in 2010, only this time it will be downward.
And even the BLS says that the birth/death numbers have little statistical meaning. The following is from their own website (courtesy of Dennis Gartman) [emphasis obviously mine]:
“Birth/death factors are a component of the not seasonally adjusted estimate and therefore are not directly comparable to the seasonally adjusted monthly changes. Instead, the birth/death factor should be assessed in the context of its effect on the not seasonally adjusted estimate… The components are not seasonally adjusted separately because they do not have particular economic meaning in and of themselves.”
Unemployment supposedly dropped last month by 0.1%, to 5%. How could a loss of jobs mean a rise in employment? Because the statistics mask a rather disturbing trend. The number of people working part-time is rising rapidly, and they are counted as employed. Again, From Philippa Dunne of The Liscio Report:
“Almost 3/4 of the gain in non-agricultural household employment [from the household survey] came from those working part-time for economic reasons, and another 83% came from what used to be called ‘willing’ part-timers. Yes, that adds to more than 100% – 154% to be precise – because fulltime employment declined by 375,000. The increase in those working part-time for economic reasons was at the 93rd percentile of all months since the series began in 1955; the decline in fulltime employment was at the 90th percentile.”
This employment report was ugly, when you look at the numbers under the headline statistics. It is no wonder consumer sentiment is down.
Dazu der Chart, der den tatsächlichen Stellenabbau (blaue Säulen) zeigt:
