Slow Growth in Productivity: What Do We Know About "Why?" and "What Is To Be Done?"—Long
J. Bradford DeLong :: U.C. Berkeley, NBER, and WCEG :: 2016-09-09 :: http://tinyurl.com/dl20160909b | https://www.brookings.edu/events/the-productivity-puzzle-how-can-we-speed-up-the-growth-of-the-economy/
The productivity puzzle: How can we speed up the growth of the economy?
Slow Growth in Productivity: What Do We Know About "Why?" and "What Is To Be Done?"—Long
J. Bradford DeLong :: U.C. Berkeley, NBER, and WCEG :: 2016-09-09 :: http://tinyurl.com/dl20160909b | https://www.brookings.edu/events/the-productivity-puzzle-how-can-we-speed-up-the-growth-of-the-economy/
The productivity puzzle: How can we speed up the growth of the economy?
What Are We Talking About Here?
- Perceived poor performance of the American economy.
- Measurement issues are fascinating and important, but largely orthogonal to perceived poor performance.
- Slow growth in overall productivity only one and not the most important aspect of perceived poor performance:
- Inequality.
- Declining mobility.
- Millennials.
- And they have, of course, been offset by the enormous decline in what can only be called caste privilege.
Eight Things Are Going On
- Always: the problems of measuring economic growth in productivity levels and living standards
- Since the 1920s: the problems of the shift to information goods—non-excludable and personalized
- And four more things:
- The post-1973 productivity slowdown
- The post-1995 productivity speedup
- The post-2004 second slowdown
- The post-2008 catastrophe of what we will soon be calling “the Longer Depression”
- Plus: reversing the causes of the slowdown may not be possible, and may not be the best policy even if it is possible
- Plus: what is it reasonable that we should expect out of the growth process?
Problem: Simply Assessing Economic Growth
- Are we, notionally:
- Measuring growth by taking today's median income, and setting it equal to that multiple of median income in the past that would provide equivalent value?
- Measuring growth by taking that fraction of today's median income that is equal to median income in the past, in the sense of providing equivalent value?
- Year-to-year changes simply do not sum or multiply
- Additional inescapable questions of moral philosophy to which “shut up and calculate!” is not a proper answer
- Narrow utilitarianism questions:
- Productivity and opportunity—I’ve never been a believer in any sort of sharp distinction between “wealth” as some form of “positive liberty” and “freedom” as some form of “negative liberty”.
- If you’re in a locked cage and lack the money to buy a key, your freedom to buy a key is illusory
- Productivity and distribution:
- Are we average people (believers in Negishi prices: that your proper weight in the social welfare function is your lifetime wealth, or perhaps your lifetime wealth squared)
- Are we median people
- Or do we aspire to be among the Ones Who Walk Away from Omelas?
- Productivity and opportunity—I’ve never been a believer in any sort of sharp distinction between “wealth” as some form of “positive liberty” and “freedom” as some form of “negative liberty”.
- Broader political economy and moral philosophy questions:
- Productivity and society—how much growth is needed and how much churn can be tolerated in order to avoid the disappointment of those reasonable expectations that keeps the altruistic-punishment mechanisms that maintain our sociability from going haywire?
- Productivity and humanity—what is the proper telos of the Sociable Language- and Tool-Using East African Plains Ape?
Problem: Valuing Information Commodities
- When you sell a rival, excludable commodity, there is producer revenue and consumer surplus:
- The first—producer revenue—is what is counted in GDP
- The second—consumer surplus—is not counted in GDP.
- When you provide a non-excludable information commodity produced with enormous amounts of non-rivalry:
- You give away the commodity
- Thus all of the final value is consumer surplus: none is producer revenue.
- You then sell, as an ancillary service, the eyeballs and eardrums of users to advertisers:
- This a much smaller money flow, a much smaller share of consumer value then producer revenue is in the case of a Smithian commodity
- Moreover, advertising expenditures are not captured anywhere in real GDP
- They are an intermediate service, not a final good.
- How large is this?
- Well, how much of their recreational time are people spending engaged with such information age commodities?
- How much of the utility they accrue during that time is due to that engagement?
- There is also the "personalization" part of the information economy: getting you the information you need to purchase exactly that monopolistically competitive variant of the commodity that pleases you the most
Problem: Valuing Information-Driven Personalization
- Henry Ford thought people got value—utility—from basic transportation:
- “Any color you want, as long as it is black…”
- Alfred P. Sloan disagreed
- GM and Chevy drank Ford’s milkshake
- Is there real value from getting people the exact monopoly-competitive variant they want?
- Or is it just a grift—selling people Chevy parts at 100% markup because they are under a Cadillac shell?
- I favor “grift”: I am more than half on Team Let’s-All-Dress-in-Identical-Blue-Overalls Myself
- But I do have three-touch smartphone advance ordering at the nearest Starbucks of a whole-milk triple-shot short latte with two pumps of sugarfree hazelnut syrup and extra cinnamon…
The Four Trend Breaks: Magnitudes
- A U.S. population come 2025 of 350M…
- Before 1973: could expect real GDP come 2025 of $44T @2009 prices: $125K per capita
- Before 1995: could expect real GDP come 2025 of $28T: $80K per capita
- Before 2004: could expect real GDP come 2025 of $32T: $91K per capita
- Before 2008: could expect real GDP come 2025 of $26T: $75K per capita
- Now: expect real GDP come 2025 of $22T: $63K per capita
Four Productivity Trend Breaks since 1973 Have Halved the Expected 2025 Productivity of the USA
- From $44T @2009 prices—$125K per capita—to $22T: $63K per capita That as, as Joe Biden would say, a BFD
- How are we to understand these four breaks in the trend of productivity growth we have seen since 1973?
The First: The Drop Post-1973
- The first of these trend breaks comes around 1973
- It is the drop from $125K to $80K to of notional 2025 GDP per capita at 2009 prices
- With Martin Neil Baily in the room, whatever summary I attempt to provide will be rightfully dismissed as cocksure and inadequate. So * I will just say: it’s complicated…
The Second: The Jump Post-1995 * The second of these trend breaks comes in 1995 * It is the jump from $80K to of notional 2025 GDP per capita at 2009 prices * Oliner and Sichel (and Fernald) and company have, I think, nailed this one to the floor… * A Fed committed to keeping deficit reduction from inducing a recession * 4% of GDP of financial flows diverted from buying government debt looking for a home * It finds a home in Silicon Valley * Just as infotech reaches a critical mass * And Intel ramps up its development pace
The Third: The Drop Post-2004
- The third of these trend breaks comes in 2004
- It is the drop from $91K to $75K of notional 2025 GDP per capita at 2009 prices
- John Fernald and company have been working full time on trying to understand it…
- I would be rash and foolish to try to challenge John Fernald and company…
- But it’s a very big shift to happen in a very short time with no obvious system-wide shock…
The Fourth: The Drop Post-2008
- The fourth of these trend breaks comes in 2008
- It is the drop from $75K to $63K of notional 2025 GDP per capita at 2009 prices
- It is a consequence of the Longer Depression
- Labor force attachment
- Worker skills
- Investment
- Experimentation with business models
The Last of These Is Robert Barro’s “Non-Covery”
- We should have had substantial bounce-back to trend after September 2009.
- We didn’t.
- It’s not because the economy is incapable of reallocating resources.
- It reallocated resources fine away from housing and into exports and investment over 2004-2008
See? Over 2006-2008 We Reallocated Resources Out of Housing on a Huge Scale
- As financial markets changed their minds about risk and return in housing…
- We moved huge amounts of resources out of the housing-construction sector…
- And, smoothly, into business investment and exports.
The Diagnoses of the Problem We Heard Here Yesterday…
- Seemed to me to focus around problems of “reallocation” and “dynamism”…
- Flawed in a relative neglect of the difference between a high-pressure and a low-pressure economic environment…
Whether “Reallocation” Is into Unemployment or a Booming Sector Is Really Important
- Keynes (1936): “I defend [my policies] … as the condition of the successful functioning of individual initiative.... If effective demand is deficient... the individual enterpriser who seeks to bring... resources into action is operating with the odds loaded against him. The game of hazard which he plays is furnished with many zeros.... The increment of the world’s wealth… [reduced] by the losses of those whose courage and initiative have not been supplemented by exceptional skill or unusual good fortune. But if effective demand is adequate, average skill and average good fortune will be enough…"
- Davis and von Wächter (2011):
- U-3<6%, mass-layoff workers lose 1.5 years' worth of pre-displacement earnings.
- U-3>8%, mass-layoff workers lose 3 years' worth of pre-displacement earnings
- Chodorow-Reich and Wieland (2016):
- An area with reallocation +1SD -> employment 2% lower at the end of a national recession-recovery cycle.
- Reallocation has no effect if it occurs during an expansion.
The Problem Comes After 2008
- The financial-crisis freeze-up of the economy…
- As housing, exports, investment collapse…
- And the Recovery Act is undersized by, arithmetically, a factor of 6
And Then We Hit the Economy on the Head with the Austerity Brick, Over and Over Again
- We also fail to restructure housing finance to encourage construction to help all those people move out of their sisters’ basements.
- And, with interest rates at zero, the Fed finds no way to signal exports and business investment to take up even more of the slack than they do.
Can We Still Recover from the Post-2008 Disaster?
- We now need to start calling it: “The Longer Depression”…
- Back in 2009 I said that we obviously will recover, easily.
- Back in 2012 I said that we could recover—DeLong and Summers—straightforwardly.
- Now?
- Gerry Friedman still thinks we could do so.
- I’m more pessimistic: but I’m not going to argue anything with Gerry until November 15
- Nevertheless: worth trying.
- We should, right now, be desperately trying to generate a higher-pressure economy if only just to see how much low-hanging fruit can be picked via easy aggregate demand-boosting measures Greenspan did it in 1995, and we all were very pleased that he did so.
Other Policies?
- Reversing the causes of the slowdown may not be possible
- Reversing the causes is unlikely to be the best policy, even if it is possible
- However, Low-hanging fruit:
- Generate a high pressure economy.
- It might work.
- And it will do good anyway
- What else?
Aside from Strive for a High-Pressure Economy and Hope, What Can We Do?
- Where is there low-hanging fruit?
- What are our value-subtracting industries?
- Finance (we pay 8% of GDP for net services that used to cost 3%)
- Health care administration (we pay 5% of GDP for services other countries pay 1% for)
- Mass incarceration (another 2% other countries do not pay)
NIMBYism
- Residential (but how mportant is it outside of SF, BOS, DC?)
- Occupational (perhaps?
More Low-Hanging Fruit
- Truman started the process of turning the federal government into an entity with a permanently very large military (for good and sufficient reasons, but still).
- Johnson set in stone the process of turning the federal government into an insurance company too.
- “An insurance company with a military”—that’s where we are.
- Is that where we should be?
All Growth Comes from “Investment” of Some Kind
- All growth comes from somewhere: it doesn’t just precipitate out of the air.
- Net private investment roughly 7% of net domestic product.
- Other net investment-like activities of equal magnitude—figure 15% of net product devoted to growth, 10% to per capita growth
- The average net rate of return on those expenditures is 20%/year.
- There are some investment and investment-like expenditures that offer very high returns indeed.
The U.S. Government Could Do Much More
- In a world that is increasingly non-Smithian, confining government to military, insurance, property rights, and contract enforcement is just silly—even adding in standard infrastructure.
- Incentivize the private creation of public and information goods:
Why is the government so good at committing itself to pay money in the event of failure—loan guarantees—but not success?
- Prizes—cf., Michael Kremer.
- Platforms: Why is it Google doing Google Books?
- Platforms: Isn’t relying on advertising to fund write-once run-everywhere value-in-scale profoundly silly?
Let’s Take the Longest Possible View…
- Pre-1800 (upper left point is 1800-50) “data”
- Periods from 1000 years down to 50
- Impressive, no? This is essentially Michael Kremer’s pre-job market paper from back when
- Almost all growth is useful ideas
- Useful ideas are non-excludable
- Thus two heads are better than one
- This is not economics: this is numerology…
- But the basic Michael Kremer two-heads-are-better-than-one model from 1992:
- Estimated over the pre-1800 period
- Admittedly with GDP rather than population on the left-hand side
- Fits our current 2000-2050 forecast
- Moreover: this is consistent with Gordon’s view that the Second Industrial Revolution was a really big deal
- 1.25%/year x 100 years of extra growth
- Exp(1.25) = 3.5
- And now that extra 3.5-fold multiplication of global GDP is over
- But, still, going forward, after the exhaustion of the Second Industrial Revolution.
- It still looks as though two heads are better than one
- If world population tops out and stabilizes at 9B in 2050
- Then will world real GDP growth stabilize at 6%/year?
- Thus we have powerful reasons to believe that economic growth is not over—that human brains will continue to be a very valuable idea-generating resource.
- Provided, of course, we do not blow up the planet.
- Or cook it.
