Tuesday, January 29, 2013

A high "drift factor"

I have trouble with data on saving and savings. I've looked at some of the datasets, looked at relations (using my old discrepancy analysis method) and not found results that I could make sense of.

The words were clarified for me by STF some time back. "Saving" happens during one year. Each year's "saving" adds to total "savings" which is an accumulation over many years. The words are very clear. But the numbers, not so much.

Krugman writes:

We all know that personal saving dropped as inequality rose; but maybe the rich were in effect having corporations save on their behalf. So look at overall private saving as a share of GDP:

Krugman's Graph

The trend before the crisis was down, not up — and that surge with the crisis clearly wasn’t driven by a surge in inequality.

The downtrend, from say 1982 to about 2008 and onset of crisis: Not only personal saving, but gross private saving "dropped as inequality rose". Saez shows inequality:

Graph #2: The Top Decile Income Share in the United States, 1917-2007
From Striking it Richer (PDF, 2009) by Emmanuel Saez
From 1978, give or take, inequality was increasing (Graph #2) while overall private saving dropped (Graph #1).

Graph #2 ends in 2007, so we can't say what happened to inequality since the crisis. However, from 1980 to the crisis, saving and inequality moved in opposite directions. And from 1980 back into the 1940s, saving increased while inequality was really quite stable. So I'm not sure that there's any correspondence at all between inequality and saving -- except that both of them show trends that changed around 1980.

But we can say that while inequality was fairly constant (1950-1978) and income was growing, saving was increasing and the MPC held good. In the more recent years when inequality was rising, Krugman's graph contradicts the MPC. So, where is the problem? what is the source of this contradiction?

I got to wondering what (if not inequality) might have caused the pattern visible in Krugman's graph. That sharp spike after about 2008, that's probably the response to the crisis -- the great concern that used to be called "panic". But set aside the crisis years. It's more instructive to look at trends of a "normal" economy.

From the late 1940s to around 1982 saving increased (relative to GDP). From 1982 to the crisis, saving fell. What explains this? My first thought was maybe "money growth" or "the growth of debt" might account for it. I know there was a slowdown in total debt growth around 1986. The last high point in the middle of Krugman's graph comes after 1982...

Then it struck me that the "up to 1982, down since 1982" pattern is similar to inflation.

Graph #3: Krugman's GPSAVE per GDP (blue) and the Rate of Inflation (red)
Yeah, I don't know. Both are quite disturbed before the early 1950s. Both rise to a mid-graph peak (but not with any great similarity that I can see). And both trend downward to the disruption of the crisis. But they just don't seem very similar.

Of course, interest rates followed a similar pattern:

Graph #4: Krugman's GPSAVE per GDP (blue) and the FEDFUNDS Interest Rate (red)
Actually, this strikes me as a better match for some reason. And I think economists would say that rising interest rates draw money to savings. That may be so -- if you have money to save. If you don't, the interest rate isn't much of an inducement. In my experience, it isn't much of an inducement. I'd rather consider Graph #3.

Could rising prices induce me to save? Probably not. But a rising income might. Even if my wages in the 1970s were not keeping up with rising prices -- and I had no way of knowing whether they were keeping up, at the time -- a growing paycheck might induce me to save. So, that would tend to produce an uptrend in "GPSAVE/GDP" when the rate of inflation was on the rise.

(It also fits exactly with the MPC. As my income grows, my spending grows (but not so fast) and my saving grows faster. Doesn't matter that my income is growing because of inflation.)

Could a slowdown in the rate of inflation cause me to save less? Again, I would think so. If my wages are not going up so fast, I might feel that I have less "extra" money, and might therefore save less. And if wages are failing to keep up with prices, well, then I can surely find things to do with my money other than save it.

What I think, in general, the thing that strikes you as the reason you do things, is probably one of the real reasons for the economic decisions people make. What we think motivates others, not so much. But what we think motivates ourselves can only be a real driving force.

So are there a lot of explanations for why Krugman's graph goes up and down?

There must be.


Luke Smith said...

Rising income and rising interest rates might influence people to say. Something interesting I heard from Michael Pettis, is that in China, when interest rates decrease people save more. He says there may be a cultural propensity to save tied into interest rates, and that people in China feel the wealth effect with rising interest rates. So, they naturally save when interest rates fall; making borrowing readily, and cheaply available.

When looking at gross savings (private and government) there is no large spike during the recession.

The Arthurian said...

Oh, great graph. It's like they belong together, GPsave and GGsave. FRED offers so much, it's good to know somebody who knows time series that I don't.

Interesting that Krugman's graph trends down from ~1982, yours from maybe 1966. Putting the two together, I see government saving was significant until the early 1970s, then not again till the Clinton years.

It needs more looking-at, I think.

Bubbles and Busts: Inequality Really Is Holding Back the Recovery said...

[...] Not convinced by Krugman’s analysis, Arthur Shipman considers other potential reasons that gross saving might have fallen in response to rising inequality. As he wisely points out, the above trend is reminiscent of the change in inflation rates [...]

While I won’t disagree with Arthur’s conclusion, I think a potentially satisfying explanation does exist if one disaggregates gross private saving.