Jon Hilsenrath Is Scratching His (And The NY Fed's) Head Over The Job Number Discrepancy And Okun's Law
A month ago Zero Hedge, based on some Goldman observations, asked a simple question: is Okun's law now terminally broken? Today, with about a one month delay, the mouthpiece of the New York Fed (which in itself is nothing but a Goldman den of central planners, and Bill Dudley and Jan Hatzius are drinking buddies), Jon Hilsenrath shows that this is just the issue bothering his FRBNY overseers. In an article in the WSJ he ruminates: "Something about the U.S. economy isn't adding up. At 8.3%, the unemployment rate has fallen 0.7 percentage point from a year earlier and is down 1.7 percentage points from a peak of 10% in October 2009. Many other measures of the job market are improving. Companies have expanded payrolls by more than 200,000 a month for the past three months, according to Labor Department data. And the number of people filing claims for government unemployment benefits has fallen. Yet the economy is barely growing. Many economists in the past few weeks have again reduced their estimates of growth. The economy by many estimates is on track to grow at an annual rate of less than 2% in the first three months of 2012. The economy expanded just 1.7% last year. And since the final months of 2009, when unemployment peaked, the economy has expanded at a pretty paltry 2.5% annual rate." Hilsenrath's rhetorical straw man: "How can an economy that is growing so slowly produce such big declines in unemployment?" The answer is simple Jon, and is another one we provided a month ago - basically the US is now effectively "printing" jobs by releasing more and more seasonally adjusted payrolls into the open, which however pay progressively less and less (see A "Quality Assessment" Of US Jobs Reveals The Ugliest Picture Yet). After all, what the media always forgets is that there is a quantity and quality component to jobs. The only one that matters in an election year, however, is the former. As for whether Okun's law is broken, we suggest that the New York Fed looks in the mirror on that one.
As for the rest, here is a guest post from early February that went over all the "outstanding" issues on Okun.
Submitted by Jeff Snider, President & CIO, Atlantic Capital Management
It's Far Deeper Than Broken Okun
ZeroHedge’s post on the apparent breakdown of Okun’s “Law” (http://www.zerohedge.com/news/okuns-law-latest-casualty-central-planning...) highlights the ongoing tragicomedy of how the science of central economic planning eventually confounds, and then consumes itself. Economics is, after all, a social “science”, an elaborate study of human beings and, most importantly, human interactions. Robert Okun, for his part, merely observed in 1962 that when “output” (whatever statistical measure is en vogue) rises by 3%, the unemployment rate seems to fall by 1%. For some reason, economics assumes that if it is true in the past, it will be true forever, so it was written into the canon of orthodox economic practice.
Economics has inferred causation into that relationship, giving it a layer of permanence that may not be warranted. Econometrics has always had this inherent flaw. The science of modern economics makes assumptions based on certain data, and then extrapolates them as if these assumptions will always and everywhere be valid. There is this non-trivial postulation that correlation equals causation. In the case of Okun’s Law, it seems fully logical that there might be causation since it makes intuitive sense – more economic activity should probably lead to more jobs, and vice versa. But to assume a two-variable approach to something that should be far more complex is more than just dangerous, it is unscientific.
In fact, Okun’s Law has already been adjusted somewhat, most famously by Ben Bernanke and Andrew Abel in their 1991 book. It was upgraded to a 2% change in output corresponding with a 1% inverse change in unemployment. Apparently with the economic “success” of that period, Okun needed a re-calibration.
Any such academic exercise means a careful review and study of the time series of economic data. Most of these academic papers focus on finding a new regression or other statistical relationship that better “fits” erstwhile independent variables into each other’s gravity. As it was with Bernanke and Abel, Okun’s law in 2000, updated by the new data set of the Great “Moderation”, needed to harmonize with the new data series of the 1980’s. Bernanke and Abel were not disagreeing with Okun, rather they were measuring some of the additional complexity that goes into the relationship between output and employment.
The academic sense of understanding the economy requires ceteris paribus. In order to make econometric models manageable and statistically meaningful, complexity has to be trimmed and managed. It is far easier to incorporate simple relationships that appear to work over specific periods of time than to try to estimate the massive and dynamic complexity that the real world exhibits. The grand mistake of economics is this sense of permanent simplicity.
In terms of broken Okun, simplicity has meant that it only measures the quantity of jobs. Okun’s practitioners assume, ceteris paribus, that a new job is a near-perfect substitute for an existing job (or a lost job from the current dislocation). In the case of job growth since the recession trough in 2010, though, “new normal” jobs are not perfect substitutes for lost bubble/artificial jobs. Newly created employment just does not produce the same wage income as the previous bubble paradigm (a lot of this disparity can be seen by the greater proportion of part-time jobs, far more than economic models predict).
This complexity and nuance extends even beyond wage income. A job in the preceding artificial period also meant easy access to cheap credit money. So job growth during the twenty plus years previous to 2007 meant both wages and available credit money (NINJA were not marginally significant until the very late bubble period). New jobs in 2010, 2011, or 2012 (and likely beyond) not only produce less wage income, there is no additional boost through credit, and therefore less marginal economic activity per new job unit. So numerically, the number of jobs, unqualified by any additional measurements, has grown without living up to the expected output growth (of course there are other factors as well, especially the negative effects of ZIRP on savers and the corporate preference for financial innovation over productive innovation, as well as the negative effects of commodity prices acting out central bank inflation expectations, but those would mean that money itself has changed in terms of how it relates to the economy – the academic sense of velocity – which should lead to all sorts of soul searching among monetary policy planners). On an apples-to-apples basis, new employment simply does not and cannot match the artificial boost of the preceding period, so a simple number quantification is less than useful.
Instead of questioning these simplified relationships and laws, mainstream economics attempts to explain all of this through a hugely negative output gap (in other words, the theories are correct, the results are flawed). Believing that past relationships still hold on largely untransformed terms, they believe that the economy should be growing much faster than it has. To make up that massive gap, the Fed embarks upon its real dual mission of making sure no large credit creators ever fail (money elasticity) while pushing investors into “risk” (and out of the supposed paradox of thrift) by making safety or saving expensive. The Fed keeps banging its head against the wall, trying greater and more costly interventions all in the vain attempt to close the theoretical output gap that does not really exist (results over theory). If economics would see that the previous paradigm of leveraged wage income is what is permanently broken (along with economic circulation through asset inflation, commonly known as the wealth effect), it might be able to conclude that the Great Moderation was really nothing more than artificial growth that won’t be returning. Then punishing the economy through intervention might be seen as counterproductive as it has really been.
If that artificial paradigm is now past tense, then so many of the “laws” upon which the social science of economics rests should also be rethought. This is particularly true since mainstream economics (especially econometrics) “grew up” during the Great Inflation and Great “Moderation” – the very periods of over-active central banking and credit production. Just because a relationship held during that specific time period does not mean it should then be extrapolated through all time. Economics is not physics, what seems to hold today may not hold tomorrow (this should be very apparent when a theory or set of beliefs fails to exhibit predictive ability). The real economic world is one marked by dynamic processes that are not well understood by academic theories wedded to their elegant, but ultimately static, mathematical constructions. Structured finance and securitizations, as an example, worked extremely well under static assumptions (such as real estate prices only move in one direction), but once the dynamic world moved beyond statistically assumed financial tolerances it was a total disaster. The paradigm shifted but the theories and causal assumptions did not until after it was all over – again, not a very scientific result.
Economics should be undergoing a more rigorous examination of its philosophical bonafides, especially since it has shown very little predictive capacity (hard science needs to be both predictive and replicable, a standard economics has not, nor will ever, meet). The foundations of mainstream economic thought should be shuddering at the prospect of being so wrong so often. Every assumption and relationship should be re-evaluated as to whether it was really true in the universal, timeless sense, or whether it was simply captured by a specific pattern in a defined and limited data series (such as the over-worn trope that low interest rates are always stimulative).
For impartial observers, there was a rather clear demarcation between the “new normal” of this recovery and the artificial bubble period, turbo-charged by trillions in securitized debt, that preceded it. Okun’s law is not broken, it was simply never a law or rule of thumb to begin with. Bernanke and Abel were probably correct that a 2% output change led to a 1% inverse change in unemployment at the time they wrote their book. The hubristic mistake of modern economics is believing Okun, and every other economic law, to be a universal property of all economic systems at all times. The world is more complex than that, and capable of changing in ways that cannot be oversimplified. Sometimes correlations are coincident to larger interactions and patterns, and not causation at all.
Real estate prices do, in fact, go down as well as up. Low interest rates do not always stimulate economic activity. Most importantly, economic potential is not simply a measure of economic output from 2003-2007. Every economic and monetary intervention (all that ails this current economic age) flows from this mistake of oversimplification and pattern bias. There would be no need to punish savers. There would be no special place afforded to the oversized behemoths of credit production and finance. Hell, there would be no place for the overgrown financial economy to begin with if economists would admit that the world does not easily fit within the Garbage In Garbage Out confines of mathematically modeled oversimplifications.
The biggest simplifying mistake in economic history was believing that a centrally planned, debt-based economy was a near-perfect substitute for a real capitalist economy, working bottom-up through unfettered price discovery, that values real production. That such soft central planning apparently worked during the Great Moderation is nothing more than pattern bias, a flawed theory captured by a unique data set of oversimplified variable relationships. The fact that it is all breaking down now is solid evidence of that paradigm shift, an invalidation of previous assumed causal relationships, not some ephemeral headwinds.
Economics is really nothing more than opinionated interpretations and analysis. But this kind of subjectivity does not jibe with central planning. Economic control is much easier to accomplish if it is conducted under the veneer of objectivity and science, especially when your marginal goal is historic impoverishment through debt. Although, that might be too harsh for most economists since they simply believed their own theories, including the idea that the marginal pace setter of artificial economic activity, asset price inflation, only moves in one direction. Their scientific observations of the Great “Moderation” told them so, and so they remained captured by their own work. The fact that an entire asset class had collapsed during all that moderation should have been a huge and unmistakable warning to economists to truly observe and learn the dangers of extrapolation and the short-comings of trying to create a science out of pattern bias and oversimplification before deliberately charting a course for historic impoverishment. But even that oversimplifies what might really be the problem here – that modern economics is not only not a science, it is an ideology.