Sunday, February 17, 2019

What happens when we can't know what we are talking about?

From What does it mean to speculate that growth is faster than we think? by Scott Sumner. There are plenty of complex system issues of which we are aware but have insufficient information to make effective decisions.

Climate change is the one I harp on with frequency - no ideological dog in this fight, just observing that climate is an extremely complex amalgamation of multiple complex dynamic, interactive, log-sensitive systems of which we have too little knowledge and radically too little information. We have very little accurate, unadjusted, consistent climate measurements before thirty years ago. A way shorter time period than the normal system cycles we know are involved in climate. And what we do have is primarily in a small handful of locations around the globe. Most the surface area is unmeasured. We are shooting in the dark.

Made worse by skewed incentive schemes and an over-reliance on system modeling which we know is prone to produce outcomes built into them through unstated assumptions. There likely is reason to be concerned but we cannot state that concern with great confidence. Instead, we skip that step and move straight towards emotional advocacy.

Sumner is talking about a similarly complex issue and possibly as consequential but with a far smaller audience of self-designated experts or interested parties - in Tyler Cowen's words:
Many people suggest that we are under-measuring the benefits of innovation, and thus real rates of economic growth are much higher than we think. That in turn means the gdp deflator is off and real rates of interest are considerably higher than we think. Someday we will all realize the truth and asset prices will adjust.

Let’s say that view is correct (not my view, by the way), how should that change your investment decisions?

[snip]

More generally, if real rates of return are high, but not perceived as high by most investors (who are still victims of fallacious “great stagnation” arguments and the like), at some point those investors will learn. With more rapid growth enriching the future, and with the realization of such, there will be a sudden demand to shift funds into the present, so as to equalize marginal utilities. So bond prices will fall and that means you should short bonds and buy puts on bonds.
It is actually a little more complex than that - are we measuring economic growth accurately, are we measuring innovation accurately, and are we measuring well-being accurately?

The answer to all three questions is almost certainly NO! but we don't have good ideas for putting ourselves into a better position to know.
Sumner adds a fourth concept that makes the questions even harder to answer.
Growth is getting increasingly hard to measure as we move from an economy of stuff (commodities) to an economy of intangibles. If we can no longer measure growth in terms of quantity of “widgets” being produced, we need some measure of the value provided by economic output. You could use money, but the value of money itself changes over time. So that won’t work.

Economists typical speak in terms of “utility”. But as far as I know there is not a shred of evidence that we have more utility than we had 60 years ago. When I look around, it doesn’t seem like people are happier than when I was a child. Maybe they are happier (I’m agnostic on the question), it just doesn’t seem that way to me. Of course utility and happiness are not necessarily the same thing, but I’d make the same argument for each. It’s not clear we are getting happier, and it’s equally unclear that we are accumulating more utility.

Now you might argue that this is because real wages have stagnated for 60 years. I don’t agree, but that argument cannot account for China, where polls show no aggregate increase in happiness since the 1990s. No one disputes that real wages in China (as conventionally measured) have soared much higher. Unlike Americans, the Chinese really do have lots more stuff; it’s not just quality change. So economic growth is a slippery concept, which is hard to measure.
Happiness? What the hell thing is that to measure?

Sumner elaborates his argument, straying, in my opinion, into some fundamentally unsound positions. But the whole piece, including the comments attached, go to the heart of a pertinent challenge. We can conceptualize more than we can effectively, consistently and reliably measure.

We get growth, innovation, and prosperity wrong and concerns about climate a century and a millennia from now become somewhat moot.

What do we do about public dialogues around issues whose dynamism and complexity outstrip our capacity to reliably know through measurement?

No comments:

Post a Comment