What is a double dip recession?

You hear the term "double dip recession" bandied about in the media a lot these days.  But there is no strict definition for what a double dip recession is or what it actually means to an economy. Have no fear; At the weekend, Robert Shiller of Irrational Exuberance fame came up with a working definition of double dips, which I will highlight and expand upon here.

Shiller’s definition of a double dip recession

Here’s what Dr. Shiller said:

I use a definition of a double-dip recession that doesn’t emphasize the short term. Instead, I see it as beginning with a recession in which unemployment rises to a high level and then falls at a disappointingly slow rate. Before employment returns to normal, there is a second recession. As long as economic recovery isn’t complete, that’s a double-dip recession, even if there are years between the declines.

Under that definition, there has been only one serious double-dip recession in the last century — and it was serious indeed. It started with the 1929-33 recession, which was followed by a recession in 1937-38. Between those declines, the unemployment rate never moved below 12.2 percent. Those two recessions, four years apart, are now typically lumped together as one event, the Great Depression.

Many negative factors persisted between those dips. High among them was a widespread sense then that something was amiss with the economy. There was a feeling of uncertainty that discouraged entrepreneurship, lending and spending, and most important, hiring.

We have to deal with a similar — though less extreme — problem today.

Most people look at the time between recessions as the only variable that distinguishes a normal recession from a double dip – and therefore concentrate on the 1980-82 period in the US as the most relevant example. However, Shiller’s definition differs from most in that it is not predicated on timing. He says a double dip recession is a period in which employment, production, income, consumption and growth dip, resume growth, and then dip again without necessarily re-attaining previous levels before the first dip began.

Shiller is pointing to a double dip in which the recession-like period I have been calling a technical recovery as a sort of interregnum between twin downturns, that never blossoms into the full-blown recovery that you and I would understand as recovery.

My definition of a double dip recession

I think it is fair to say that Shiller’s description of a double dip recession is rooted in a sense that the underlying causes of the first and second dip are the same – that the recession, technical recovery, renewed recession really feel almost like one big long recession.

This is what I have been saying for months now. In fact, I believe the recession dating committee has stalled in dating the end of recession for exactly this reason. Witness my comments in this regard from last January:

There has been a lot of conflicting evidenceregarding this fake, stimulus-induced recovery we are now witnessing.  NBER committee member Robert Gordon of Northwestern University made statements this past Spring suggesting he sees an early recovery dating (see Jobless claims may signal the end is near from April 2009). But Martin Feldstein, another NBER dating committee member, is holding out…

But, without wading into the debate on multipliers (you can see my piece on multipliers here), I think we are likely to see long enough a period of stimulus-induced growth to force the dating committee into a late summer 2009 demarcation for the end of the recession. Read my piece on recovery in my recent post Readers of this blog expect the recession to last redux when I said the technical recovery’s start would probably end up being dated Summer 2009.  If I had to pick a month, I would go with August which is a month before I had been predicting the recession would end early last year.

–Is the recession dating committee preparing for a double dip?

But, more than that, I think Shiller’s comments about the Great Depression point to structural issues. We are in the midst of significant secular developments regarding employment, credit and debt right now – much as we were in the 1929-1938 time period and in the 1970-1976 time frame. I have said that I consider the twin recessions during those episodes to be double dip recessions. They represent two long depressions, one deflationary and the other inflationary.

Near the end of the recession dating committee post, I wrote:

Question: are the 1970 and 1973-75 recessions or the 1929-33 and 1937-38 recessions double dips as well?  I think so. That’s why I spoke of a depression as a “decade is one of sub-par global growth with short business cycles punctuated by fits of recession” in a March 2008 post. In my view, the discussion about multipliers is less relevant.  It is the D-process of deleveraging and debt deflation leading to depression which dominates the secular trend. Any credit guy like me will tell you the balance sheet always trumps the income statement.

So, I think that’s where we are: in the technical recovery phase of a double dip recession that is a once in a generation period of balance sheet repair. If you want to use the Great Depression as a guide, we are closer to 1931 in terms of events and timing (see Thinking about Creditanstalt today). But the policy response has been much more aggressive this time around.  Ultimately, I think that means a technical recovery has taken form, even if the underlying fundamentals are suspect.

My hope is for a multi-year recovery – which was my prior baseline before the chances of a double dip increased. A multi-year recovery really should be the baseline outcome in any technical recovery scenario. However, this does not rule out a double dip per Shiller’s definition. In my view, this journey is far from over – as the continued low savings rate in the US and the policy errors on both sides of the Atlantic demonstrate. The shift, like those two previous periods, will take at least a decade to complete.