viernes, 22 de marzo de 2013

viernes, marzo 22, 2013


Will the Real Unemployed Please Raise Your Hands?

By John Mauldin

Mar 20, 2013



This week’s letter will be a very short part of a book I am writing with Bill Dunkelberg (the Chief Economist of the National Federation of Independent Businesses) on the future of employment. It has taken longer to write than I initially anticipated, for a host of reasons, chief among which is that the future is not as obvious as I originally thought. Diving into the data has brought a few surprises. It doesn’t help that I have (probably to the frustration of Dunk, although he is way too polite to say it) changed the focus from “merely what we need to do to create jobs (which is still an important part of the book) to what kinds of jobs will the future bring and who will get them.


But to understand the future of employment, we have to be able to measure what we mean when we say employment. And the data that we all too often think of as hard and fast is anything but. Is unemployment in the eye of the beholder? We know what we mean when we say our brother-in-law is unemployed. But does the government data mean the same thing? The answer is “maybe, sometimes, and it depends.”


I selected this part of the book not only because it is toward the beginning but also as a result of a conversation I had this week that got me to thinking about data and its usefulness. It is in this context that we will look at unemployment data.


But first, a quick comment on why this letter is not about Cyprus, which seems to be the topic du jour. I wrote four weeks ago, after my visit to Athens, that Cyprus would be a problem and to pay attention.



Can Someone Figure Out Cyprus?



The story changes every few hours. It is not clear what the Cypriot parliament will do. They could quite possibly simply say no, or change any number of things. The whole thing is patently ham-handed. It illustrates what I mean when I keep saying that the EU is making up the rules as they go along. Why take money from widows and then exempt the branches of Greek banks? (I hear the reasoning, but I don’t understand it.) It is all about protecting banks and institutions and not the defenseless. I find the whole thing rather outrageous in the way that it looks through the wrong end of the telescope.


And dear gods, why risk creating a situation that encourages depositors in banks to question the ownership of their money? If the deposits of the citizens of a member country of the EU are not safe from bureaucrats, at least up to €100,000, then why should they trust any bank? And why in Hades is the IMF appearing to dictate that depositors will lose money? And why do Turkish depositors in non-Greek banks lose when Greek depositors in Greek banks don’t? Cyprus is complicated enough without poking that nationalistic anthill.


Cyprus is also a problem in itself. It is the size of a small city, not a country. My friend Frank Trotter of Everbank noted today that the agricultural subsidies Texas gets from the US government total more than the entire GDP of Cyprus. Would anyone blink if the banks in San Bernadino County went bankrupt? Cyprus’s banking system is around 8 times the size of the country’s GDP, so asking the country to back the deposits of its banks is ludicrous. Perhaps as much as half the deposit total is from outside the country (encouraged by the banking system and regulators), and it is widely assumed that Cyprus launders a great deal of Russian-oligarch and Russian-mafia money. Which of course does not sit well with Germany, especially in advance of its upcoming election. Try telling the good burghers of Bavaria that they should pay to bail out Russian oligarchs. Frau Merkel has made it clear she will not.


The EU contends that Greece was a one-off situation and no one else will get bailed out, at least not on sovereign debt. We will see.


While the bailout money needed is about the size of the entire economy (€17 billion), that is chump change to maintain the system. The banks are in large part bankrupt because they bought too much Greek debt and were forced to take haircuts on that debt. If the ECB bailed out Greek banks and depositors (which they did), then why is Cyprus a special case, demanding that some small depositors and not others lose money? Then again, if the banks go down, which they could easily do without a bailout, depositors stand to lose a great deal more, as Cyprus as a country can no longer access the bond market.


Does Europe want Russia to step in at the price of Russia’s getting a Mediterranean port? Is saddling one set of people with a tax and letting others off the hook even legal in Europe? There seems to be some debate.


This mess occurred because there is no clear eurozone banking policy or general deposit insurance, both of which were promised in the wake of the last such crisis and then soundly rejected by Germany (Merkel), which does not want to pay for the banking sins of other countries.


I fear that whatever I write will be both obsolete and wrong before I can even hit the send button, so I will forbear. We will revisit Cyprus when we can make some informed observations. But let me point out that I wrote a few weeks ago that the real challenge to the euro is, first, France (another country whose banks are far too large to be bailed out in a crisis) and, second, that the voters of more than one country are simply getting fed up. The way European leaders are handling the Cyprus situation does not inspire confidence.



Data, Information, and Opinion



As I wrote in last week’s Outside the Box, I visited Bill Bonner’s rather isolated hacienda in the Andes, where he lives two months of the year. (His ranch has expanded to some 500,000 acres, by the way.) Bill is the founder of Agora, now the largest investment-letter publisher in the world, but he started from scratch some 30 years ago in Baltimore. As the rest of the group was out riding horses, Bill and I stayed back, waiting for our turn; and we began to reminisce about old times. The first time I met Bill was back in 1982, when I went to Baltimore. The city had sold him two buildings for $100, and he had paid too much. To say it was a dangerous neighborhood back then is an understatement.


As a rather naive young Texan, I was scared to get out of my car and walk a half block. (For good reason, I later found out.) But the price was right and the low overhead was an advantage to a young entrepreneur in an industry that was still just a hope and a dream in some of our eyes. We had lots of hard work and hard-won knowledge ahead of us!


We talked about the evolution of the newsletter business. We both started in the days of printed letters and large promotional mailings (remember those?) And we were both lucky enough to transition to what was to us a very new and unknown thing called the internet early enough to get some advantage. But some things stay the same. We are still dealers in information, trying to help readers understand what is going on and helping them with their investments.


There is a hierarchy to what is offered in publications these days. The first and lowest level is data, the raw numbers. On top of that come the levels of information, news, analysis, opinion, and actual actionable advice. Each adds to our depth and breadth of understanding. Yet, even as they add to our store of knowledge, we find ourselves drowning in information and knowledge. All too much of it is random noise, serving only to drown out clarity and wisdom.


The highest form of writing, I said to Bill, is storytelling, and he agreed. (Bill is the master of the story in our business.) Both of us, with our Irish ancestry and proximity to the Blarney Stone, come naturally to the storytelling medium. In an earlier time, we might have been fixtures at the local pub, spinning our tales for a pint of Guinness. But the fates have been kind to us, and we get to earn more than a few pints here and there.


Our goal (and that of many of our friends) is not to be the teller of old tales, but to seek to find the analogy, the resemblance of what is happening in the world to what we can readily understand. If the reader can come away with a fresh insight into the mysteries of life and markets, of our larger social contracts and how they impact our financial future, then we have done our work.


Our craft is part of the human heritage, harking back to when our ancestors sat around the fire at night relating the events of the day to the larger picture. (I once had a very-late-night discussion with Anatole Kaletsky about whether economists were not in fact modern-day shamans; but rather than discerning the future in the entrails of sheep, we deduced certainty from government data. I mused that some economists might be better off peering at intestines.)


Scientists have found that we humans actually get an endorphin rush when we arrive at an understanding of an event that that has perplexed us. Sadly, that understanding does not have to be true, but merely believed, to release that ancient chemical rush.


The problem is that we writers often have to challenge accepted wisdom. The “facts” are often slippery things, and it is our job to corral them carefully, lest we and our readers be led on a horseback ride into the high Andes of faulty assumptions.


And thus we come to employment. Each month in the US, on the first Friday of the month, we breathlessly await the release of the employment numbers. Markets move on the whisper of a trend, only to reverse when that whisper turns out to be just one more bit of random noise. It is said that one should not want to know how sausages or laws are made. I would add government statistics based on surveys to that list. But let’s now turn to the book Dunk and I are writing on employment, and take a look at our first draft of the chapter on government employment data.



Will the Real Unemployed Please Raise Your Hands?



The old saying about liars, damn liars, and statistics contains more than a grain of truth. Statistics are especially hard to swallow when we don’t like what they tell us. How many times do we question the validity of some statistic when it disagrees with our preconceived notions? (Statistical surveys say, a lot!) And unemployment numbers for the last few years haven’t given us reason to celebrate.


Let’s start with what you already know: millions of people are unemployed. How do you know this? The same way you know how many Americans believe UFOs are real, or how many watch the Super Bowl or own a foreign car. Almost all such data comes from surveys.


Surveys can be wonderful tools, giving us insights into data our minds can’t otherwise process. What we have to remember is that every survey is based on a sample. There is no name-by-name database (as far as we are aware) containing information on all our UFO beliefs. If there were it would be outdated almost instantly, as every minute thousands of people are born, die, enter the country, leave the country, change their minds, or get abducted and leave the planet.


Most national surveys use a sample of, at most, a few thousand people. From this, we extrapolate conclusions about a country of 300 million people. Real-world experience indicates this is usually sufficient, too – or at least close enough for most purposes.


Survey respondents can only answer the questions they are asked, however, so it’s important to ask the right thing in the right way. Otherwise the results won’t reveal the information you want to know.
BLS: Everyone’s Favorite (Whipping Boy) Agency


Officialinformation about jobs, or lack thereof, comes from the Bureau of Labor Statistics, which is part of the US Department of Labor. BLS gathers, analyzes, and publishes a staggering amount of labor-related data. The bls.gov website is a treasure trove for people who are interested in employment trends.


For the purposes of trying to understand what the BLS does, let’s look at their “Employment Situationreport. The one we will consider emerged from the BLS womb on Friday, January 4, 2013, at precisely 8:30 in the morning, Washington time. The 41-page news release summarizes the survey data BLS gathered the prior month. A new report is born each month, typically on the first non-holiday Friday. It is the source of the “unemployment rate reported in the media and closely scrutinized by economists.


The report began with the two key numbers:

The monthly change in nonfarm payroll employment, and

The unemployment rate

This particular month those numbers were 155,000 and 7.8 percent, respectively. The first number is supposed to reveal how many new jobs the economy generated that month. The second is the percentage of the workforce that was unemployed during the same month.


Figure 1: BLS News Release

How does BLS know? The numbers come out of not one but two different surveys the agency conducts every month.


The Establishment Survey asks a sample of employers how many people work for them. BLS adds up their answers and extrapolates the results to the whole economy. In other words, US employers collectively had 155,000 more people working for them in December than they did in November. Or so BLS believes.


A separate Household Survey calls on peoplenot businesses – to determine their employment status. The result this time is that 7.8 percent of respondents who “wanted” to beemployed” were notemployed,” as BLS defines those words.


Now, why does it take the BLS 41 pages to convey what we just said in a single paragraph? Details, details. Let’s consider a few before we move on.


Seasonal Adjustments: Unlike people, all months are not created equal. Some of the events that affect employment are unpredictable, like earthquakes. Others are semi-predictable, like hurricanes and blizzards, but don’t happen every year. And some are very reliable.


We know, for instance, that in the summer students are out of school and families tend to go on vacation. We know people do more shopping in December. We know car makers introduce new models in September, and that they order parts months before then. All these events happen every year, more or less.


BLS statisticians consider these factors and makeseasonal adjustments” to their numbers. This helps make each month’s report comparable with other months. However, since not everyone wants these adjustments, or may disagree with the methodology behind them, BLS publishes the unadjusted numbers, too.


Other, less frequent adjustments, such as new census data compiled every ten years or demographic changes like the Baby Boom help account for longer-term changes. These adjustments are not an exact science, but without them the data would eventually stop telling us anything useful.



Labor Force: An unemployment rate of 7.8 percent begs the question: 7.8 percent of what? You might think the answer iseveryone.” Not quite. Not everyone is able to work. Not everyone wants to work, either. The unemployment rate is a percentage of a subset of the population BLS calls the “labor force.”


In fact, the labor force is a subset of a subset. Unlike the US Census, which counts every human being it can find, BLS is interested only in the “civilian noninstitutional population.” People in prison may be chipping rocks or making license plates, but BLS doesn’t care. They are “institutional” and not competing for jobs (unless you are a license plate maker or a rock chipper, perhaps). People in hospitals, nursing homes, and various other categories are also excluded.



Figure 2: Household Survey, December 2012


As we see in Figure 2, in December 2012 the civilian noninstitutional population was 244,350,000. The civilian labor force was 155,211,000.


Now, how does BLS determine how many Americans are in the labor force? They call us, or at least enough of us to provide a statistically valid sample.



The Household Survey



Formally known as the “Current Population Survey,” the BLS Household Survey is a monthly sampling of about 60,000 households. It is actually conducted by the US Census Bureau, which then gives the raw data to BLS for further analysis.


The 60,000 households are called during a “reference week,” normally around the 12th of the month. Whoever answers the phone is asked a series of questions about their work status. From their answers, BLS classifies each person age 16 or over in the sampled household into one of three categories:


Employed

Unemployed

Not in the labor force.

Every person who lives in a sampled household goes in one of these three groups. The only escape is to be in the “institutionalizedcategory, as we discussed earlier, or to be under age 16.


BLS considers a person to beemployed” if they were a paid employee at any point during the reference week. If you started a new job the day before, worked fifteen minutes, and then got fired, BLS still calls you employed.


People also count as employed if they worked in their own business or farm, or if they worked without pay at least 15 hours in a family business or farm. This is important. Suppose you were laid off last month, and instead of looking for a new job you decided to start your own business. This is a fairly common scenario during a recession. Even if no one has hired you, your start-up work means you are “employed.”


Also counted as employed are people who have a job but who were temporarily absent during the reference week because of illness, bad weather, vacation, a labor-management dispute, or for other personal reasons.


The BLS sample, when extrapolated to the entire nation, tells us the US had 143,305,000employedpeople in mid-December 2012. To be classified as “unemployed,” on the other hand, people in the sampled households must meet all of the following criteria:


Had no employment during the reference week,

Were available for work the entire week, and

Made specific efforts to find employment sometime in the last four weeks.

(Note that the BLS unemployment data has nothing to do with unemployment benefits. That’s an important piece of information, too, but it’s not part of the BLS report.)


Under these definitions, BLS believed there to be 12,206,000 unemployed people in the US in December 2012. Add this to the 143,305,000 who were “employed,” and we get 155,511,000 either working or who want to work. This is the “civilian labor force.”


From here the math leads straight to the unemployment rate: 12,206,000 divided by 155,511,000 equals 7.8 percent. A year earlier it was 8.5 percent, so maybe we’re making progress – but don’t be too sure.


The “civilian noninstitutional population” is 244,350,000 people. Subtract the labor force, and we have a third category that BLS callsnot in the labor force.” At 88,839,000, it is a big group, all of whom are neitheremployed” nor “unemployed”; they simply don’tparticipate” in the labor market.


BLS tracks this group with a statistic called “participation rate.” They found that 63.6% of the population was either working or looking for work in December 2012.


The participation rate has been drifting downward for a long time, and the reasons are a big subject of debate. Most analysts agree that at least some nonparticipants could be in the labor force if they would onlyparticipate.


Recall that you are notunemployed unless you looked for a job in the last four weeks. You can be physically able to work, and even willing to work if someone offered you a job, but you must still have madespecific efforts” to find a job in the last four weeks. If you didn’t, then BLS does not consider you unemployed.


Nonparticipants aren’t necessarily watching TV all day. At any given time, many people are in various kinds of in-between conditions.
For example, someone might leave a job (or be fired) and spend a few weeks or months traveling, or caring for a sick relative, or engaged in volunteer projects.


Obviously, they would need to live on savings or find some other means of support, but many people do so.


A more ominous example of a nonparticipant is the “discouraged worker.” Such people had a job, lost it, could not find another one, and are not making the “specific efforts” to find employment that BLS wants to see.


They tend to start looking again when they think more jobs have become available. This leads to a seeming paradox in which the economy improves, enticing discouraged workers to look for jobs, which expands the workforce but makes the unemployment rate worse instead of better.


Say what? Yes, the unemployment rate can change sharply, even if no one is hired or fired, based simply on changes in the participation rate. In the December 2012 example above, we saw that 88,839,000 people were “not in labor force.” Suppose 1% of this number had instead been looking for work that month. Not finding work, just looking. That would be 888,390 people now counted as officially unemployed.


Look what happens. The number of unemployed people goes up from 12,206,000 to 13,094,390. The labor force is now 156,399,390 instead of 155,511,000. The unemployment rate isn’t 7.8 percent, it’s 8.4 percent. And what if not just 1% but 5% of thosenot in labor forcestart makingspecific efforts” to find a job? Unemployment rises to almost 11%! Other surveys have suggested that 5% is not at all an unreasonable assumption as to how many “nonparticipants” are actually seeking employment, and that number might even be way too low.


This kind of massive change is entirely possible with the BLS methodology. Does it mean they are doing it wrong? No, it simply means they are trying to distill a complex situation into a comprehensible set of numbers. Distortions are inevitable.


What it really tells us is to take all the numbers with a pinch of salt. Political spin and media reporting are not much help, either. When the unemployment rate goes from 7.8 to 7.6, they all exclaim, it’s a huge improvement, we’re on the right course, we’re making progress! And a move the same size in the other direction is a sure sign of recession and the loss of hope. (Even including a number after the decimal point in the unemployment rate demonstrates that economists have a sense of humor, if somewhat distorted.)


The BLS, ever-helpful, also gives us other ways to measure employment, but they are largely ignored by the mainstream media because they paint a much bleaker picture. (The following note is from my friend Grant Williams, in his brilliant Things That Make You Go Hmmm...)


The fact that there are multiple unemployment rates, including one that is labeled as the 'official' rate (U3), should be enough to raise a red flag as to how easy manipulation of those figures can beparticularly as the most comprehensive unemployment rate (U6) is almost double the official unemployment rate. It includes


part-time workers who want to work full-time but cannot due to economic reasons

'discouraged workers', or those who have stopped looking for work because current economic conditions make them believe that no work is available for them and,

other 'marginally attached workers', or 'loosely attached workers', and those who 'would like to' and are able to work but have not looked for work recently.


And, as can be clearly seen from the tables below, this has been the case for a long, long time.




Source: Portalseven


Instead of fixating on any one month, we should all look at long-term trends. The various statistical anomalies tend to sort themselves out, given enough time. The unemployment numbers are repeatedly revised for several months, then revised annually, and even larger adjustments are made every few years, based on new and more accurate data. For instance, two months later, we find that the number of new jobs in December 2012 has risen to 219,000. That is almost 40% more than the original estimate. But the January 2013 number has been revised down by 38,000. Such large revisions are typical.


Remember the jobless recovery of the Bush years? Revisions made years later tell us that it was not a jobless recovery after all. The seasonal adjustment factors are based in large part on recent trends. If the trend has been down, the seasonal adjustment is likely to understate the number of jobs created.


Likewise, the unemployment numbers in the recent credit crisis have been revised downward, as the employment trend prior to the crisis had been upward.


The BLS has to have some mechanism by which to make its estimates. They really do try to get it right, but they are working with imperfect data. The formulas they use are quite public, if a bit arcane. Contrary to the belief of some, there is nospin” in their data. It is what it is. But the formulas they use can be quite controversial to those who care about such things and are hotly debated (although hotly might be a bit over the top as a descriptor of econometric disputes).


Further, the BLS has to simply guess each month as to the number of new businesses that have been created and the number of businesses that have failed. The ratio of the two is known as the “birth/death ratio,” and it can make a big difference. The “realbirth/death number may not be really known for years after the initial report, until other sources of information (tax data, etc.) can be collected and analyzed.


This brings up two points to always keep in mind. First, it is generally the trend of the BLS revisions that is more important than just the revealed numbers. In general, the trend can actually give us a lot more useful information than the fresh monthly data. If you add the trend data to the information from other surveys and sources, it can give a hint as to the direction of the economy and the markets.


Second, anyone actually trading on the monthly employment data when it is released deserves whatever revenge the markets take. (If your investment advisor is prone to such nonsense, you might consider what else he or she is up to, and question whether the relationship is really contributing to your well-being. Just saying.)


With those caveats, let’s answer the question posed as we began this chapter. How bad is the jobs crisis? BLS data gives us at least a rough idea. For comparison, we’ll go back five years, to December 2007. We now know that the recession was already unfolding, but at the time economic confidence was still generally high.



                                                          December 2007                     December 2012
 
Monthly Change                                      +18,000                           +155,000

Unemployment Rate                                  5.0%                                  7.8%

Civilian Noninst. Population             233,156,000                         244,350,000

Civilian Labor Force                         153,866,000                          155,511,000

Employed                                          146,211,000                          143,305,000

Unemployed                                          7,655,000                            12,206,000

Not in Labor Force                              79,290,000                            88,839,000

Participation Rate                                 66.0%                                        63.6%



We could write an entire book exploring just the numbers in this one table. A few highlights and anomalies:


The unemployment rate was much better in 2007 better, in fact, than the 5.5% some economists now considerfull employment.”


The civilian noninstitutional population grew approximately 11 million in five years, yet the civilian labor force rose by only 1.6 million.


The number of employed was actually 1.9 million lower in 2012 than in 2007.


The “not in labor forcecategory gained 8.5 million people in five years.


Anyone who grumbled at the December 2007 BLS report would probably be glad to have it back. The situation now is much, much worse. It is also worse for some than others. But therein lies a book, and it is time to draw this letter to a close.


Have a great week. Learn to ignore the noise, unless it is the gentle cry of a new granddaughter. And then just enjoy it.


Your wondering what my grandkids will do when they grow up analyst,
 

John Mauldin

Copyright 2013 John Mauldin. All Rights Reserved.

0 comments:

Publicar un comentario