There is no precise definition in economics of a recession. The rule-of-thumb definition, "two consecutive quarters of negative GDP," sounds good until you realize that GDP is measured in quarters, not months, and there is more than one way to measure GDP, and that GDP is only one measure of overall economic activity.
This is why the NBER uses the term "downturn" because in economic reality, the period of falling economic activity is likely surrounded by difficult times; either before or, more generally, after. The downturn is one specific part of an economic cycle: the part where final economic activity (that is what GDP measures) is falling in inflation adjusted terms. That sounds like a mouthful, because it is.
When measuring economic output, suppose I make a bolt, and I sell it to you to make a motor, you sell it to someone else to make a wood chipper, and someone else buys it to run a business. If you count each sale, it distorts the economy, because it would seem like a larger economy than if someone sat down and made the wood chipper all in one shop. So, GDP is a measure of the final sale. This, and other murky realities of figuring out what is going on, fall under the rubric of econometrics: how to measure what we have measured.
The NBER often shrouds what they do in a great deal of mystery, but it isn’t as hard as all of the conference calls and arguments make out. The NBER tends to date downturns since the Second World War by periods of rapidly rising Unemployment Rate. The headline unemployment rate is a mixture of two things: demand for jobs, and demand for labor. It is a poor measure of the supply of either, since the job market is very inefficient and opaque. Anyone who believes that all information is transmitted instantly in a market needs to get an email asking you to do a job you did 15 years ago now and again.
Now the demand for jobs and the demand for work are going to cross at some point, that is where the market is. If all workers who want jobs at prevailing wages can find them, and all employers who want workers can find them by offering the prevailing wage, then the market is said to "clear." This isn’t the usual circumstance in the job market, for a long list of reasons. The two pieces are also impure: there are people who want to work, and there are people who need to work. Marginal theory in economics tells us that the first person to give up working, is the person who feels it is no longer worth their trouble; for example, the time and expense of searching for a job. Searching for a job is as expensive as having one, because, well, searching for a job is your job.
The same is true on the other side: businesses give up on hiring when it isn’t worth the trouble to hire the worker.
This means that job demand is influenced not only by need, but by factors such as inflation. High inflation drives people to work who otherwise might not, and stagnation causes people to drop out of working. As someone who has been job hunting for the last few months, I can tell you that the job market is not only bad, but it is loaded with perverse incentives for employers and employees alike. For employers there are two points: one is for inexperienced people who are cheap now but will grow into the role. They need someone long term. The other is for avoiding the learning curve for some urgent need right now. Hence, if one does an analysis of job postings in a variety of areas, and that is what I did, job requirements are either much longer, with exacting lists of skills, or much shorter. Either the employer needs the person yesterday, or they need them next year. Then employers aren’t looking for job seekers, but job switchers.
This is a long introduction to looking at two sets of numbers which came out recently: a preliminary GDP, and July’s employment report. The Administration touted the GDP figures as proof that the stimulus plan is working. However, the internals of these numbers under cuts that assertion. First, Defense has shown itself to be the most efficient GDP use of fiscal policy. This makes a kind of perverse sense: defense demand was new actual demand. Most of the "tax cuts," and much of the "construction spending," as well as a good deal of the UI went to savings directly or indirectly. States didn’t do new construction, they cut construction to preserve tax cuts. So if Obama wants to call his stimulus plan "escalate in Afghanistan," then yes, his fiscal policy is working.
The other part which undercuts Obama is the Unemployment Rate. Remember that unemployment is demand for jobs and demand for workers. People left the work force "in droves," as Forbes magazine observed. If we were seeing stimulus, then people would be entering, not leaving, the workforce. Longterm unemployment would not be creeping up either. Instead, what we are seeing is a deflationary pressure. People are giving up looking for work and getting by on what they have. Which is to say, they aren’t.
So let’s get back around to the relationship between the "downturn" and unemployment. When the unemployment rate stops moving up quickly, it is a good sign that two things have happened. One is that businesses are no longer contracting, and the other is that people are no longer being driven into the job market by inflation. Ergo, from the policymaker’s standpoint, inflation is no longer a risk, and the bleeding has stabilized. It is time for a different policy mix. That’s why UI, not GDP, is the better predictor of when the NBER will declare a downturn over.
But most of us are not policymakers inside Washington DC, or CFOs of companies with large policy exposure. Most of us are job seekers, and for us, there was no good news. Obama was not talking to us about the economy mending, he was talking to investors: "please buy our private debt so we can stop borrowing public debt" is the message he was sending.
For us the more relevant prediction comes from numbers uber-guru Mark Zandi, who sees this as a stabilizing economy; but that the job market, that is what the rest of us live in, won’t recover until next year. The downturn is over for Wall Street, but not for Main Street. The story the numbers tell is this: about two years ago, a real estate bubble popped. The people who were overleveraged in that bubble were the first to get hit; but when they went down, the people who were overleveraged in people who were overleveraged went down harder: the lack of regulation made the system wired like a Christmas tree and ready to explode. Policymakers met this mild beginning of the downturn with an apathetic sort of fiscal stimulus package — too little too late. They met the inflation wave with a botched strong dollar play that ended up taking out even more overleveraged overleveragers. When this happened, the dominoes fell fast and basically the next step was that the world’s rich went on strike. Instead of a depression created by little people doing a run on the bank and striking against poor wages, we had the wealthy do a walk on the banks and go on a buying strike.
The world economy went into free fall. That free fall is conclusively over, and the Administration is talking up the post-free fall economy. They have to, because there isn’t much they are going to actually do about it. The moves by the Obama administration were to cajole the global investing class to come back to the table. By and large they have: we are now back to where we were before the global elites went on strike.
However, even taking out the fiscal crisis, we are staring at a larger downturn than America has seen since 1981-82. That’s right, take out the free fall and this is worse than the last two downturns.
Which gets me to the graph above. What it is, for those who have not seen it before, is the percentage of payrolls lost from the peak of employment, and how long it takes to recover. You can see the deep valley of 1981-82, the hammer blows of the late 1950′s recession and the 1948 recession. You can see this being worse than any single downturn is now all but a done deal.
What is crucial is two things. One is on the chart, you can see how the 1990′s downturn and the 2001 downturn took forever to get out of. If this downturn takes as long, we will never recover from it. Before the jobs return, inflation in resources will force another round of contraction. Oil is at 70 dollars a barrel; we have, roughly, about 50 dollars a barrel of increase before New Economic Geography runs in reverse on the globalized supply chain: business becomes unprofitable to run, production starts to relocalize, and capital is stranded. The other is something that is obscured by this chart: that is economic crises comes in clusters. There have been three consensus changing periods of economic change: the demobilization from World War II, which crested in the 1948 recession. This heralded a period of liberal consensus, which the Republicans attempted to take control of. However, the next crisis was on their watch, the late 1950′s early 1960′s pair of employment downturns. If you want to know why Kennedy was so popular, it is that he really did get America moving again. The last was 1975-1982, a period of 7 years with three recessions, two of them very sharp.
The present downturn is close to over, but the recovery from it is crucial. In most post-War recessions, as you can see, we had a dramatic rebound in hiring. Once the disinflationary pressure was removed, the economy bounced back. However, the last two downturns have been more like depressions than recessions: they were driven by deflation, and hiring took much longer to come back. Insufficient inflationary pressure lengthens the hiring crisis.
The job figures tell us that that pressure, the deflationary one, is still in place. The GDP figures say that the economy is not surging, even though oil prices already are marching upwards. The collision of the two — one says inflate more, drive more money in; the other says pull back, slow down — is the next wave of challenge. It will be longer, more difficult, and more complex. The first six months of Obama’s policy, good and bad, will be swamped by how the more difficult transition of the American economy from being oil consumption constrained to something else is accomplished.