Response to Marc Andreessen on Secular Stagnation | Hacker News Reader

Response to Marc Andreessen on Secular Stagnation

larrysummers.com

43 pointsmr_golyadkin8 comments

From a non-economist point of view, the bit about deflationary technological advancement doesn't seem hard to understand at all.

Summers claims it makes no sense because price changes are accounted for in GDP. I'm willing to bet that GDP is not adjusted for Moore's Law, or we would have had much higher official growth numbers than we've had. Every iPhone is a more powerful computer than an early-90s supercomputer. Each iPhone sale does not add as much to GDP as each supercomputer did, even after adjusting for our modest inflation.

And yet that computational power is improving people's lives in all sorts of ways.

There's usually a paragraph in most introductory macroeconomic textbooks, as they introduce the definitions of nominal GDP and real GDP, to the effect of "Real GDP is GDP deflated by the price of a basket of goods. It cannot account for qualitative improvements in the nature of goods, nor can it account for substitution effects where an old good is completely replaced by a new good of equal price but higher quality. These effects are difficult to measure and difficult to quantify." And then because they are difficult to measure and difficult to quantify, they are ignored in the rest of the textbook.

It's telling that a professional economist like Larry Summers can forget that paragraph. When you work with models day in and day out for your whole career, the models quickly become reality, even if the models have known flaws and blind spots that become very apparent when you apply them to everyone else's reality.

I still remember the first day in my college freshman year Econ 101 class. The professor began by explaining how the foundational idea of the discipline is that people are rational actors who make decisions by weighing costs and benefits. I thought of all the actual people I knew, every one of whom had made major life decisions based on impulse, emotion or stupidity, and thought "welp, so much for economics then."

I think the decline in real interest rates is explained by the market's coordination toward systemic risk collusion:

Now that we know that governments bail out inadequately hedged systemic risk, we might as well all play a new game in which we ignore those risks entirely. Perhaps in the past we under-hedged, but now we don't have to hedge at all.

Original tweet storm can be found here: http://www.yvoschaap.com/pmarca/

"From an economist’s point of view, this paragraph is very hard to understand. Real GDP and productivity statistics are calculated after adjusting for price changes – so they are unaffected by inflation or deflation."

Glad to hear that Mr. Summers has had a change of heart on deflation being bad for real GDP. Progress!

> they are unaffected by inflation or deflation

In other words someone picks a basket of goods and those price changes are measured.

Economists have every incentive to get this right. I think they probably do, but if you see the world progressing in a "singulatarian" way toward exponential growth in many areas, it's unlikely that the basket of goods chosen accurately tracks the broader trend.

Also, if such a singulatarian world is happening, economists, by virtue of not believing that to be the case, would correct for those goods when formulating the basket used to track price changes.

One final point: There are enough seemingly price-stable goods where the prices are actually the result of factors other than pure supply and demand (such as unknown risk premiums, etc.) which if they end up in the basket of goods could cause systemic bias, since those risk premiums may not be structurally stable in either direction.

From a policy perspective, we must recall that interest rates are a blunt instrument and it has been a major goal/concern of the Fed to come up with other, smaller knobs to turn. One of these is QE policy, others are regulation and enforcement mechanisms that impact capital flows in a broad way (but less broadly than interest rates).

Reply on news.ycombinator.com