BASED ON DATA RELEASED LAST WEEK, Ben Bernanke’s track record on forecasting growth in gross domestic product is two-for-two — an impressive showing for his first year as Federal Reserve chairman. But since he’s zero-for-two in forecasting the unemployment rate, it does make you wonder if he has any misgivings about holding the line on the short-term interest rate. Does he now secretly believe he would have been wiser to raise it a notch or two?

This scorecard on Bernanke is not based on any forecast he made recently, but on the forecasts all Fed chairmen make in testimony before committees of the House and Senate in February and July of each year. Since that testimony accompanies the Federal Reserve Board’s “Semiannual Monetary Policy Report to the Congress,” the projections represent his team’s most visible effort to get the outlook right.

The need to get the outlook right was emphasized by Bernanke himself, in his own testimony of July ‘06. Because of the “lags between policy actions and their effects,” he declared, “we must be forward-looking, basing our policy choices on the longer-term outlook for both inflation and economic growth.”

Similar views were expressed repeatedly by Ben Bernanke’s illustrious predecessor, Alan Greenspan. Yet based on his own testimony before Congress in February and July, Greenspan’s track record in forecasting economic growth was poor by any standard. So far, at least, Bernanke has done much better.

The forecast given for growth of real GDP is measured as the percentage change in fourth-quarter GDP from the fourth quarter of the previous year.

We can judge Greenspan’s track record according to the easiest standard possible — the “range” rather than the more narrow “central tendency” — and still find it lacking. In the last 12 years of his chairmanship — from 1994 through 2005 — his February forecast turned out to be right only once. That is, only once did the actual rate of GDP growth come in within the range given.

As you might expect, however, his July forecast proved to be more accurate, since by then the fourth quarter is only three months away. Over the same 12 years, his July forecast turned out to be right twice. That is, only twice did the actual rate of GDP growth come in within the range given.

By contrast, Chairman Bernanke not only got GDP growth right in both February and July of his very first year; he also got it right according to a more stringent standard. Based on data released last week, we now know that in fourth-quarter 2006, real gross domestic product ran 3.4% higher than a year ago. And in both February and July, that fell within the range of the central tendency offered by Bernanke.

But if he was two-for-two on GDP growth, he was zero-for-two on the unemployment rate. As of both February and July, his central tendency for the unemployment rate through the fourth quarter was 4¾% to 5%. In fact, the unemployment rate averaged 4.5% in the fourth quarter. And if even if you tacked on the January reading of 4.6%, released Friday, you’d still average 4.5% from November through January.

In other words, the Fed chairman underestimated the tightness of the labor market — a mistake that sounds eerily familiar. Through the late 1990s, his predecessor made the same error and was eventually forced to raise the short-term interest rate too far and too fast. True, Bernanke can point to signs that wage inflation still looks relatively tame. But he is not the sort of Fed chairman who prefers to wait until he sees the whites of inflation’s eyes.

The upside surprise of fourth-quarter GDP growth might be having a delayed impact on him. When he delivered his July testimony, he had just finished hiking the interest rate target on federal funds to 5.25%. And when the Federal Open Market Committee decided to keep the fed-funds rate at 5.25% at its Dec. 12 meeting, it erroneously stated in its press release: “Economic growth has slowed over the course of the year.”

At its meeting last week, on Jan. 31, the FOMC again decided to keep fed funds at 5.25%. But that was only shortly after it learned that, far from slowing, economic growth had accelerated in the fourth quarter. This time, the FOMC release acknowledged: “Recent indicators have suggested somewhat firmer economic growth.” In fact, that “firmer economic growth” was the only reason the Fed chairman’s July and February forecasts for GDP growth proved correct.

So we might imagine the Fed chairman rereading his testimony with some satisfaction. My guess is that, over the next few months, he’ll be keeping close watch on the unemployment rate. If it doesn’t rise to that central tendency of 4¾% to 5%, he may start tightening fed funds to nudge it up there.

You can now purchase my book t.v. talk from the CSPAN website. It was an informative discussion and elaboration of the topics covered in my book, and includes a question and answer session.

http://www.c-spanstore.org/shop/index.phpmain_page=product_video_info&products_id=194920-1

RUMORS OF THE LABOR MARKET’S IMMINENT swoon proved much exaggerated. Wednesday, the Employer Services division of Automated Data Processing declared that, based on its monthly survey, the private sector had shed jobs in December, the first such decline since April 2003.

But bears impressed by this widely followed survey were in for a rude awakening.

Friday, the far more authoritative jobs report from the Bureau of Labor Statistics showed just the opposite. The private sector added 150,000 jobs in December, while total jobs on nonfarm payrolls expanded by 167,000. With upward revisions to previous months, payroll employment in December ran 1.4% higher than a year ago, well within the range of 12-month gains over the past two years. The unemployment rate held steady at 4.5%, close to its five-and-a-half year low.

So the labor markets ended the year on an upbeat note. Bears who predict the Federal Reserve will have to ease the short-term interest rate some time this winter began to cast their eyes on spring.

Rumors of the labor market’s long-term swoon are also much exaggerated. Take, for example, the myth of the understated unemployment rate.

The fall in the labor-force participation rate — the share of the eligible population either working or looking for work — has spawned a cottage industry of critics who cite it as “proof” that the unemployment rate is no longer a reliable measure of either the economy’s ability to use its workers or workers’ ability to get jobs. The participation rate ran 66.2% in 2006, down from its peak of 67.1% from 1997 through 2000.

Critics figure this extra 1% that used to participate in the labor force would be doing so if only the jobs were available. From there it is an easy leap to say that the unemployment rate is seriously understating the extent of unemployment.

But the critics ignore overwhelming evidence to the contrary. In fact, the BLS has kept comprehensive data on the “hidden unemployed” that are meant to address this very issue. Are people not looking for work for “reasons of discouragement”? Are they not looking for work because they might have difficulty getting child-care or adequate transportation? Are there people who work part time and therefore would be counted in the labor force, but would like full-time work and can’t get it (the “involuntary part time”)?

The agency uses these categories to supplement its count of the official unemployed. People are considered unemployed if they did not work in the past week and made at least some effort to look work over the past month. But the hidden unemployed have always been present in large numbers — even in 2000, when the official unemployment rate was at a 30-year low, and jobs were going begging. Were they present in unusual numbers over the past few years? No.

The BLS keeps three other measures of the unemployment rate that incorporate the “discouraged,” “marginally attached,” and the “involuntary part time.” All three confirm that hidden unemployment has been no greater over the past few years than it was in the late-Nineties.

Speaking of the unemployment rate, I spoke about it at a presentation I did in November about my book, Econospinning, taped by Book TV. If you’re otherwise unemployed (and sleepless) at midnight EST this Saturday, Jan. 6, you can view it on C-Span2. For times in your area, log on to www.booktv.org.

For exact times in your local area, log on to:
www.booktv.org/schedule/

I’ll be interviewed, live, on “The Larry Parks Show” this coming Thursday, December 14th between 7:00 and 7:30 pm on New York City cable-t.v., Time Warner Channel 56 and RCN Channel 84. Be sure to tune in!

-Gene

SOME NUMBERS ARE PSYCHOLOGICALLY IMPORTANT, especially when given analytical significance they do not quite deserve. Friday the Institute for Supply Management announced that its closely watched index of manufacturing activity — which generates one such number — unexpectedly fell below 50 for the first time in years, to 49.5 in November from 51.2 in October. According to the institute itself, readings above 50 on its Purchasing Managers Index indicate expansion in manufacturing, while readings below 50 indicate contraction. So we are supposed to believe the dividing line was crossed. And since the manufacturing sector has entered a recession, can the rest of the economy be far behind?

What Paul Samuelson said about the stock market, however, can pretty much be applied to the PMI: It seems to have predicted about four of the last one recessions. Worse, it slipped below 50 in May 1995, June ‘98 and November ‘02 — all periods that preceded expansions.

Even a PMI of 49.5 reflects a mildly expansionary trend in the manufacturing sector. The real dividing line on the index between expansion and contraction is closer to 47. The PMI is a composite of indexes, which in turn are based on questionnaires. While the questionnaires are scored by computer to a standard that makes 50 the neutral number, it doesn’t follow that the results have to work that way. To the contrary, that’s not the way it works, as a number of economists have long been pointing out.

Tuesday Federal Reserve Chairman Ben S. Bernanke made his own statement about the economic outlook. While Fed chairmen often prove no better at forecasting than the rest of us, Bernanke this time was probably as good or better. “Over the next year or so,” he declared, “the economy appears likely to expand at a moderate rate, close to or modestly below the economy’s long-run sustainable pace.”

The chairman then confessed to harboring “substantial uncertainties” around “this baseline forecast.” If anything, we may be in for some upside surprises. As Bernanke pointed out, “the solid rate of job growth, the decline in the unemployment rate and the healthy pace of capital investment could be signals that underlying economic fundamentals are stronger than generally recognized.”

The manufacturing PMI will probably rebound above 50 in the next few months. And Friday, when the employment report for November gets released, we’ll probably find the job market still remains strong.

WHY TALK ABOUT LOU DOBBS in a column on the economy? Not because the CNN newsman has anything to contribute to our understanding of such important issues as immigration, outsourcing and international trade. Even so, Dobbs has come out with two books on these and related topics, and his nightly news show, Lou Dobbs Tonight, has an audience of nearly a million. I devoted a chapter to him in my own book, Econospinning (2006), part of which was excerpted in Barron’s (Aug. 14).

Normally, that should been more than enough ink to spill on what I perceive to be his dishonesty and capacity for hate-mongering. But the size of his audience alone suggests a need to continue setting the record straight.

In his 2004 book, Exporting America, Dobbs published a table on future job growth that blatantly misrepresented its source, the Bureau of Labor Statistics. When I pointed out the error in a phone conversation, he clearly was not pleased. Now he not only perpetuates the misrepresentation in his new book, War on the Middle Class, by rerunning the table, but this time does not even mention the source.

Just what does this table show, and why is it so important? In his earlier book, subtitled, Why Corporate Greed Is Shipping American Jobs Overseas, Dobbs left the impression that a BLS study in February 2004 found that “seven of the 10 biggest areas of job growth were in menial or low-paying service jobs” — just the sort of finding that buttressed his thesis and enraged foes of outsourcing and globalization. But the list he printed, beginning with “waiters and waitresses,” had been gleaned from another table entirely, in the back of the BLS study, and then scrambled for heightened effect.

In fact, the main finding of the BLS study was exactly the reverse of what Dobbs claimed. The study prominently stated that “employment in professional and related occupations is projected to grow the fastest and add more workers (6.5 million) than any other major group.”

In his new book, this time in a chapter called “Exporting America,” Dobbs observes that “no one is saying which industry will be the source of replacements for…jobs lost to outsourcing.” The folks at the BLS were “saying” it not only in ‘04 but in an update in November 2005, notwithstanding Dobbs’ refusal to listen. After declaring that “the following jobs will have the greatest demand for workers in the next decade,” Dobbs reprinted the same list, merely reinforcing, to his fans, the idea that outsourcing, and all it represents, had done America wrong.

Dobbs’ new book also answers another question: why he and other newscasters habitually refer to China as “communist” China — as in a Dec. 8, 2005, broadcast, in which he announced, “Still ahead…communist China planning a massive new assault on American consumers.”

The precise nature of this fearsome attack: The Chinese government had brought about 100 companies to a New York City trade fair to market their goods.

The term “communist,” together with phrases like “massive new assault,” is clearly meant to allude to a time when China was in the enemy camp — where Dobbs presumably hopes to place it once again. Yet, in response to what he calls “frequent” questions about his word choice in his broadcasts, he writes, “China is the world’s largest communist country. Its government is communist. President Hu and Premier Wen Jiabao are not elected officials. China is not a democracy; it is a communist state. The real question is, why do other news organizations and our political leaders ignore that reality?”

That circular statement deserves to be aired on Comedy Central, perhaps on The Colbert Report. Otherwise, to refute Dobbs can feel like striking a fly with a mallet. While China is hardly anyone’s model of a free country, that’s not the point. It’s just a little harder to instill fear in television audiences with talk of “communist” India or “communist” Indonesia — two countries, by the way, that the 2006 Index of Economic Freedom, released jointly by The Wall Street Journal and Heritage Foundation, cited as less free economically than post-Mao China. Besides, if “communist” means the absence of democratic elections — which it does not — then more than half the world’s nations must also be communist.

Sure, outsourcing has its drawbacks, but it’s time to call Dobbs’ economic demagogy what it is.

Dear Donald Luskin:   I write you, somewhat belatedly, in response to comments about me on your blog, www.poorandstupid.com, in the hope of clearing up certain misunderstandings between us.

   First, I want to apologize for leaving a false impression about your critical coverage of Paul Krugman.

   Then I’ll point out that you could not possibly think my own critiques of Krugman come down to “picayune nitpicking,” since you yourself have raised some of the very same issues (and even once endorsed an earlier version of one of these critiques in my Barron’s column).

   Finally, I’ll show that when you endorsed both Krugman’s and Brad DeLong’s response to my criticism of Krugman’s misuse of the employment data (Chapter 2 of my book), you made the elementary—and for you, rather shocking–error of assuming Krugman and DeLong were actually responding to my criticism.

   If I really were trying to read Krugman’s mind about where he might have gotten a number, Krugman would have been right to point out that this attempt of mine was doomed to fail. If I really were trying to criticize what “Paul” intended to write in that column—rather than what he actually wrote—DeLong would have been right to point out that what Krugman intended was above reproach. These are classic “straw men”–“an argument set up so as to be easily refuted”—and they serve a classic purpose: to avoid having to deal with argument I did make.

    First, my apology.

    There is an endnote reference to you in my book that you wrongly complain about not finding in my book’s index. But it’s not an indication that the book is “sloppily indexed”—only that my indexer followed the normal convention of omitting endnote references. Incidentally, I’m glad about that, because your main complaint about the reference to you is quite legitimate. It does leave the false impression that I dismiss your criticisms of Krugman as lacking in substance. 

    Here, in fact, is an earlier draft of what I had written, with omitted words in italics:

   “Those who love Krugman for his Bush bashing  should not take his word on the economy, any more than those who hate him for it should reject everything he says on that topic. In the latter group, while National Review on-line columnist Donald Luskin has produced some valuable criticism of Krugman’s economic reportage, he unfortunately makes it seem as though distaste for that reportage must split along party lines (not true in my case, since I voted for Bush’s Democratic challengers both times).”

   Since that second sentence seemed a bit wordy, I unwisely shortened it to the way it reads now–“In the latter group, National Review on-line columnist Donald Luskin unfortunately makes it seem as though distaste for Krugman’s economic reportage must split along party lines….”

   Again, that leaves the impression I dismiss your critiques as lacking in substance, which I certain do not. Future printings of my book will run the restored sentence (with “National Review Online” written correctly). You might still think your writings on Krugman do not sound pro-Bush, but then we’d be disagreeing on something that is fairly subjective.

   You were right on the objective fact that you do make substantive arguments. And again, please accept my apology.

   (As for your point that, by voting for “Bush’s Democratic challengers both times,” I “voted twice for idiots,” I also think you’re right. Whenever I vote in a national election for one of the major party candidates, I vote for the lesser–idiot.)  

   But your dismissal of my critiques of Krugman as “picayune nitpicking” can only be attributed to a fit of annoyance on your part.

    For example, you yourself wrote recently (Aug 4 ’06), “When the unemployment rate started falling immediately after the passage of the 2003 tax cuts, the liberal media kept repeating that it was all a fake…”

   Indeed, Krugman kept repeating that very thing about the unemployment rate by making specious arguments about long-term joblessness, phantom dropouts, the labor force participation rate, and the employment-population ratio—topics that I deal with in Chapters 4 though 7 of my book. If you’d only take the trouble to read these chapters, I’m sure you’ll find them to be the very opposite of “nitpicking.” (Also see a recent column of mine—“Working the Numbers,” Oct. 23–reprinted on this blog, about a similar attempt by a New York Times reporter to dismiss the huge gap in unemployment between the U.S. and the European Union.) And as noted, you yourself once endorsed an earlier version one of these critiques in my Barron’s column.

   I also assume you would not call the scary prospective cost of eldercare a “picayune” topic. Try Chapter 1 of my book on Krugman’s coverage of this topic—a pattern of denial that occasionally involved brazen misrepresentation (see especially pp. 10-14).

   Next consider your surprising endorsement of Krugman’s way of defending his own misuse of the employment data. You wrote, “Krugman throws out an unsourced number — that 140,000 new jobs per month are required to keep up with labor force growth –but it’s sheerly Epstein’s supposition that Krugman got that number somehow from one or the other of the DOL’s [Department of Labor’s] survey’s.”

   But I was hardly supposing any such thing. For all I know, Krugman could get an unsourced number from virtually any place—the New York Times, The American Economic Review, or one of his graduate assistants—and I wouldn’t suppose I knew that source, since I make no claim to being clairvoyant.

     Here is what I actually wrote in my book’s first reference to the “140,000” (p. 18): “He [Krugman] threw in a statement about ‘employment’ having to ‘rise by about 140,000 a mount just to keep up with a growing population’ that referred to the Household Survey data” (italics added).

         In other words, the statement about the 140,000 referred to this particular primary source known as Household Survey data. I could hardly know where Krugman himself had gotten the number from. Lest anyone think I thought I knew, I added later (p. 21), “When Krugman wrote that ‘employment…must rise by about 140,000 a month just to keep up with a growing population,’ he was using, unwittingly or not, all of the above concepts [from the Household Survey]”—italics added.

         And lest anyone still think I thought I knew, I added later, “Krugman, remember, denied that this statement had anything to do with the Household Survey. So we can only assume he got it from another unnamed source…”

         Far from supposing I could tell that Krugman had gotten this number from the Household Survey, then, I accepted his denial at face value.

         I also cannot imagine you yourself could doubt that Krugman’s statement referred to concepts from the Household Survey. Just for starters, I pointed out (p. 29) that the reference to the “growing population” hardly meant “just any population, but the population eligible for employment according to the Household Survey definition.”

         As for DeLong, please do me the courtesy of reading “DeLong Follies #4” on this blog, in which I explain how the 140,000 does indeed refer to the Household Survey. I honestly doubt DeLong was legitimately hung up on the semantic confusion over whether a figure like 140,000 could “come from” the Household Survey in the same way that other numbers Krugman cited in the column “came from” the Establishment Survey. But even if he was confused about this initially, he could not have been once he actually read what I wrote, since I was careful to distinguish between the numbers that were “drawn appropriately from the ‘Establishment Survey’ data and the 140,000 that “referred to the Household Survey data.”

   Then do me the courtesy of read “DeLong Follies #3” on this blog, in which I make the simple distinction between what DeLong says “Paul” intended to write and what he (“Paul”) actually wrote about these numbers. I keep asking DeLong to “do” Krugman’s “math” to show why the math isn’t nonsense, and for obvious reasons, he keeps ducking that daunting assignment. But unlike DeLong, you have no stake in denying the nonsense of Krugman’s math. If you still have questions about it, then read my response (also on this blog) to Wharton Research Data Service Director Michael Boldin, who tried vainly to defend the math as a kind of reduction ad absurdum on a wrong-headed way to specify “full employment” that Krugman attributed to President Bush’s “own economists.”

    Notice I point out to Boldin that what Krugman attributed to these economists was completely false. They did not even mention “full employment,” either directly or indirectly. So the math is still nonsense.

   Which raises a question: If Krugman was attributing this to the wrong source, what source did he get it from? Or did he just make it up? To answer that question, you’d have to be able to read Krugman’s mind. And you and I would hardly presume to do that. 
   Sincerely,

   Gene Epstein

My correspondent who had tried defending Krugman’s math (see DeLong Follies #7) turned out to be Wharton Research Data Services Director Michael Boldin. Following is a long e-mail I sent economist Boldin in response to recent correspondence from him plus a brief phone conversation.

Working the Numbers 

“IT AIN’T WHAT YOU DON’T KNOW that gets you into trouble,” Mark Twain once famously observed. “It’s what you know for sure that just ain’t so.”

What people know that isn’t so about the economy has often brought trouble on us all, and is the subject of my recent book….

The longest section in the book is devoted to data on jobs and joblessness in the U.S. economy. But recently, exactly the same sort of econospinning (see Chapter 7) has been applied to another consequential issue: the gap in performance between the labor markets of Europe and the U.S.

The economies of the European Union (EU) have often been touted as “capitalism with a human face.” But over the past decade, the unemployment rate has given much of Europe a most inhuman face, indeed.

Take the EU’s two largest economies, those of Germany and France. In Germany, unemployment has been running at more than 10% since 2003. In France, it only recently fell below 9% for the first time in six years. In the U.S., by contrast, the unemployment rate is now less than 5%, down from 2003 highs of a little over 6%.

If this stark difference were reversed — if the U.S. suffered these frightening rates of joblessness instead of the other way around — we can bet that much would be made of it. Instead we used to hear that the numbers were not comparable, even though conscious efforts have been made by labor statisticians to make them so.

Recently, however, the pattern of denial has given way to an even bolder response. The unemployment rate, we are now told, should no longer be the main indicator for making these comparisons. It should make room for the “employment rate” — “A Statistic That Shortens the Distance to Europe,” according to the headline in the Sept. 30 edition of The New York Times.

That “Europe has high unemployment, brought on by tight labor laws,” and that the U.S. “has low unemployment thanks to its vigorous and flexible economy…has long been the consensus view,” the article informs us. “But there is another rate that can be considered — the employment rate — and that shows that the differences are narrowing, if not vanishing, for those in the prime working ages.”

Just for starters, the reference to “prime working ages” (25 to 54), already reveals a certain desperation to slice-and-dice the data to get the desired result. With improved health and longer life expectancy, working Americans in their late-50s and early-60s might still be considered in their “prime.” Compare ages 25 to 64, in fact, and the U.S. would still be ahead, even according to this misleading indicator.

And it is misleading. About a century ago, there came into being the idea of the labor force, which consists of those who have jobs (the employed) and those who don’t, but actively seek work (the unemployed). To get the unemployment rate, you take the unemployed as a share of the labor force. If the employment rate were calculated in the same way, there would be no objecting to it — although it would then be just the other side of the same statistical coin.

In fact, the employment rate is more accurately called the “employment-population ratio” — the share of the adult population that is employed. No society has the obligation to provide work to those who don’t seek it — nor the right to do so, slavery having been abolished a long time ago. Besides, plenty of folks who are not in the labor force are quite productive, performing valuable work that just happens to go unpaid.

As a measure of how well an economy provides jobs to those who want them, the employment-population ratio might be “considered.” But it should be rejected as any kind of useful alternative to the unemployment rate.

The risks that the U.S. labor market would ever emulate Europe’s have happily faded. The hope is that Europe will emulate ours. But if the focus remains on the employment rate instead of the unemployment rate, that hope also will fade.

Brad DeLong still refuses to “do” Krugman’s “math.” (See DeLong Follies #6 and #3.)  Recently, however, “eaglesrfree2003″ wrote me to take up this challenge.
But while eaglesrfree at least did me the honor of actually reading what Krugman wrote–a challenge that still eludes DeLong–he did not bother to read what I wrote in either Chapter 2 of my book or in DeLong Follies 3. Otherwise, he would not have trotted out this version of “an equivalent Krugman-esque statement”:
Given that the gap between current employment levels and the full employment is 7 million [wrote eaglesrfree], where full employment is defined by a situation where 67 % percent of the US adult population is employed, then  it might seem that if the economy continues to produce 288,000 payroll additions per month it would take only 2 years to reach full employment.  However, population growth will expand the full employment target during this time and thus the time taken to reach a situation of full employment  will take almost 4 years, even if payroll additions average 288,000 per month.
 

But as I point out in the book (pp. 25-26), if payrolls expand by 288,000 per month, and if “employment…must rise by about 140,000 a month just to keep up with a growing population,” that leaves 148,000 jobs per month that can only be filled by the unemployed. And since there were 8.2 million unemployed at the time Krugman wrote, if you do this math, you’d find that in “almost 4 years,” there would be only a little over a million unemployed left, for an unemployment rate of less than 1%! “No economist would expect that kind of ‘full employment,’” I wrote (p, 26), “not even in his wildest fantasies.”
I’m sure Professor DeLong would understand this, if he could only be aroused from hibernation.