Saturday, March 16, 2013

Things get worse

Marcus Nunes shows this one

Graph #1: NGDP, 1992-2012, Source: Marcus Nunes, Historinhas

with a dividing line there, indicating the start of Bernanke's Chairmanship of the Fed. He wants you to see the big dip after Bernanke took over. Marcus likes to think that everything was fine until Bernanke took the chair in 2006. I disagree.

I see two dips on Marcus's graph; I see progressive decline.

Marcus always says the Great Moderation is unrelated to the Great Recession. He thinks the years before 2006 and the years since 2006 are unrelated. I think what happened after 2006 was largely a consequence of what happened before.

Marcus's graph goes back only to 1992. Extend it back a few years more,

Graph #2: NGDP, 1985-2012
and you can see another decent size dip during the 1991 recession -- as deep as the 2001 dip, yes, but only half as wide. Only half as long, timewise, the 1991 dip. Only half as much volume of lost output, give or take. The 2008 dip is twice as deep as the 2001 dip. The 2001 dip is twice as wide as the 1991 dip. And -- well look at that! -- the 1991 dip is twice as deep as the 1986 dip. Things get worse.

Off the graph to the left, the blue line was higher and the dips showing less moderation. But you know, the graph shows actual GDP -- actual-price, inflating-price GDP. So lots of the downtrend since the 1970s can rightly be attributed to the beating down of inflation. Maybe we should be looking at inflation-adjusted GDP instead.

Since 1980:

Graph #3: RGDP, 1980-2012
Consistently higher before 2000 than after. Things get worse.

And again, inflation-adjusted GDP -- this time since 1947:

Graph #4: RGDP, 1947-2012

A lot less "moderation" in the left half than in the right. But it looks to me like mostly downtrend, through the whole graph. I don't just see trouble only after 2006.

Here's another view of it -- this time not inflation-adjusted GDP, but inflation-adjusted potential GDP. Best-case, perfect-world GDP:

Graph #5: PGDP, 1949-2012

I don't just see trouble after 2006. I see long-term decline.

Oh -- and if Graph #5 shows hypothetical decline, #6 shows a real-world decline in our use of the productive capacity at our disposal:

Graph #6: Capacity Utilization, 1967-2013
Capacity Utilization shows a long-running downtrend. Things get worse.

But you know what? All the above graphs look at "percent change from year ago" values. They all show that annual growth rates are in a long-running downtrend. Let's stop and look at things a different way.

Graph #7 below shows inflation-adjusted GDP (blue) and Potential GDP (red) in billions of dollars. The lines go up, meaning GDP is always getting bigger as the years go by. Always, except after 2006 there where the blue line drops. That's the part that bothers Marcus. Bothers a lot of people, really, and for good reason.

Still, the general trend of both lines on this graph is up, not down. That's true. But what the other graphs tell me is that these lines have been going up slow and slower as time goes by. That's my point, really. After 2006, the blue line drops? Yeah, it's a problem. But I think we could have seen it coming for a long time. Because the lines have been going up slow and slower for a long time.

A lot of people might say they knew it was coming. That wouldn't surprise me at all.

What the earlier graphs show is that the lines are going up slow and slower. Doesn't look like it on Graph #7, does it? Looks like the red and blue lines are gradually curving upward. But that's a feature of growth. A mathematical property of growth. If something grows by doubling every year, it goes from 1 to 2 to 4 to 8 to 16 to 32 and before you know it, the numbers are too big to do in your head. That's a constant rate of growth, that doubling every year. But even though the rate is constant, if you plotted the numbers, you would draw a line that curves up. And it wouldn't be a gradual curve, either.

Even if the growth is much slower that "doubling every year", if the growth rate is constant, then the line curves up when plotted.

And even if the growth is slowing, rather than constant, there is a good chance the line will curve up when you plot the thing. But just because the line curves mostly up, it doesn't mean everything is rosy. Like I said, a lot of people would say they knew trouble was brewing.

If you want to look at a constant growth rate on a graph, and you don't want to be deceived by the up-curving properties of growth, then you can use a log scale on the vertical axis. When you use a log scale, a constant growth rate looks like a straight line when you plot it, not a line that is curving up.

So if you plot the thing on a log scale and the line is straight, then the growth rate is constant. If the line curves up even though it's on a log scale, then the growth rate is fast and faster. If the line curves down, the growth rate is slow and slower.

Inflation-adjusted GDP and potential GDP, in billions:

Graph #7: HOVER Graph showing RGDP and PGDP

Move the mouse away from Graph #7 to see the normal plot. Hover the mouse over Graph #7 to see the log scale plot. The normal plot shows growing GDP. The log plot shows that growth has been slowing for the last 60 years. With the mouse over the graph, the graph shows less and less up-slope as time goes by. Slow and slower growth. The problem goes back to a time long before 2006.

Park your mouse on Graph #7.

The graph shows nothing but a slowing of growth. Things get worse.


João Marcus said...

You are mixing together long term growth questions with fluctuations.
I was only concerned with economic stability, i.e. reduced cyclical fluctuations.
They are different objects.

Valeria said...

This is cool!