Friday, September 7, 2012

Shall we wonder why?


Which increased faster, prices or output?

Graph #1: Quarterly FRED data for GDPC1 and CPIAUCSL

The numbers on the vertical scale don't mean anything, far as I can tell. With a different base year for Real GDP (or for CPI) those numbers would be different. What is significant on this graph is the uptrend, the downtrend, and the flat trend.

From 1947 to 1966, real output increased faster than prices. From 1969 to 1982, prices increased faster. And since 1983, prices have been increasing almost exactly as fast as real output.

1969-to-1982 is no surprise. That's part of the time called "the great inflation". What's most interesting, I think, is the difference between the years before the great inflation, and the years after it.

A quick glance at the numbers suggests that before the great inflation, real output increased about 44% faster than prices did. After the great inflation, real output increased about 2% slower than prices.

I think this is a remarkable picture.

5 comments:

Jazzbumpa said...

Hold on a minute.

Real GDP is nominal GDP/GDef, where Gdef is the implicit GDP deflator.

GDef is not identical to CPI, but they are close relatives.

What you're doing is taking nominal GDP and adjusting it for inflation twice.

I'm not able to get my head around what that actually means. But I think you have wandered into angels on pinheads territory.

Proceed with extreme caution.

Cheers!
JzB

The Arthurian said...

Not at all. The first division, the GDP/GDef you point out, produces a new series that we may think of as 'real' output.

The second division is simply a comparison of real output and prices.

I know: division is division. However, unless I'm a pinhead, my arithmetic is valid.

Jazzbumpa said...

Oh, the arithmetic is valid. Numbers are just numbers, and you can operate on them in any way your imagination concocts.

High marks for creativity.

It's the concept behind the arithmetic that is boggling me.

The time series of the ratio of (inflation adjusted output) / (prices)
is still
(real output) / (inflation)^2.

That's the math.

[Granted the two inflation measures aren't identical, but bear with me, that's only a detail.]

Construct a narrative to help me understand its significance. What real world phenomenon does this number help us understand? What policy recommendations does it suggest? Something along those lines. Otherwise its just an experimental curiosity (not that there's anything wrong with that.)

Cheers!
JzB

Jazzbumpa said...

Correction:

The time series of the ratio of (inflation adjusted output) / (prices)
is still
(MEASURED output) / (inflation)^2.

JzB

The Arthurian said...

"The time series of the ratio of (inflation adjusted output) / (prices)
is still
(MEASURED output) / (inflation)^2."

Yep. Sort of like
(distance) / (time)^2

This question is valid: Which increased faster -- inflation-adjusted GDP, or prices?

The graph answers that question.

"Construct a narrative to help me understand its significance. What real world phenomenon does this number help us understand?"

The narrative is this: "From 1947 to 1966, real output increased faster than prices. From 1969 to 1982, prices increased faster. And since 1983, prices have been increasing almost exactly as fast as real output."