Automation: Labor augmenting or substituting? Yes.

One of the biggest questions concerning automation/robots/artificial intelligence is whether it augments human labor or is a substitute. A recent “Open Letter on the Digital Economy” outlined a research agenda that called for “(i)dentifying business models in which technology is a complement to – not a substitute for – labor and creating a taxonomy of their common characteristics.”
But the issue is not an either/or. The classic case example is where the same technology augments parts of a job while substituting for others. In the past this has been the case for machines and physical labor. A backhoe allows one worker to dig a ditch faster than a person with a shovel, even if a worker with a shovel may be needed to finish up.
More recently, technology has been substituting and augmenting knowledge work. For example, Institutional Investor recently pointed out that “RIAs Shouldn’t Fear the Robots” [RIAs are Registered Investment Advisers.] So-called “robo” advisers can handle routine transactions like portfolio rebalancing, freeing humans for more high touch activities such as discussing changes to the client’s situation.
More interesting is when a technology in one phase of development might augment a job but substitute for it later on. This appears to be the case with autonomous vehicles. A New York Times story earlier this year on Volvo’s pilot assist technology explains how the technology currently works:

After a driver pressed a button on the steering wheel, sensors scanned the road and locked on to the vehicle a few car lengths ahead. A white icon lit up on the dashboard, and the wheel began moving on its own.
As the road curved, the Volvo steered itself through it, automatically adjusting the throttle and steering. The vehicle seamlessly kept on going, though after about five seconds, a subtle dashboard light asked the driver to keep a gentle touch on the wheel.
Not that it was needed — the Volvo could keep going hands-free for miles at speeds up to 30 miles per hour on a properly marked road. But for now Volvo has programmed the XC90 to start slowing down if a driver does not heed the warning light, making the vehicle a bridge between “lane keeping” and the truly hands-free technology set to hit the market soon.
“This is about making the tedious parts of people’s drives less stressful,” said Jim Nichols, a spokesman for Volvo. “We’re not talking about a driver simply checking out and not paying attention.”

Wired ran a similar story on “The World’s First Self-driving Semi-truck Hits the Road”:

The Freightliner Inspiration offers a rather limited version of autonomy: It will take control only on the highway, maintaining a safe distance from other vehicles and staying in its lane. It won’t pass slower vehicles on its own. If the truck encounters a situation it can’t confidently handle, like heavy snow that covers lane lines, it will alert the human that it’s time for him to take over, via beeps and icons in the dashboard. If the driver doesn’t respond within about five seconds, the truck will slow down gradually, then stop.

But, as today’s Wall Street Journal (“Truckers Gain an Automated Assist”) points out,

Manufacturers consider these systems a crucial primer for developing demand for automated vehicles . . .
Eventually, they could lead to trucks that drive themselves entirely.

That might not be the end of truck drivers, however. A Washington Post story on “How self-driving tractor-trailers may reinvent what it means to be a truck driver” points out an alternative:

. . . trucking companies could be expected to find ways to turn their cabs into mobile offices for drivers.

This is an example of a third possibility, what I call “transformative”. Some technologies will go beyond either eliminating the need for humans to undertake certain activities or helping them do those activities better. They will transform the activity completely. These are the truly disruptive technologies.
To illustrate what I mean, let’s take a non-human example: the horse. In a recent article, Erik Brynjolfsson and Andrew McAfee ask a provocative question: Will Humans Go the Way of Horses? Remember, however, that horses still exist in the economy. According to one study, there were 9.2 million horse in the United States in 2003. That is a sharp decline from the estimated 21.5 million in 1900. But it is not elimination. What happen was a transformation in the horses’ role from providing work energy to recreation. It anticipate that the same transformation will occur for human drivers. Driving a car will become recreation not transportation.
What does this mean for other types of human activity? I don’t know exactly. I do know that there will be technological substitution and displacement, technological augmentation of human activities, and technologies that will disrupt and transform human activities. We need to prepare our society from all three.

July employment in tangible and intangible industries

Today’s employment numbers from the BLS show continued job growth with payrolls up by 215,000 in July. This was only slightly below economists’ forecast of 225,000. Last month’s number was also revised upward to 231,000. The unemployment rate was steady at 5.3%.
In a reversal of earlier trends, employment in tangible producing industries grew faster than in intangible producing industries. Employment in tangible producing industries was up by 126,500 in May. Trade, Transportation & Utilities and Accommodation & Food Service were once again the biggest gainers. Intangible producing industries added 88,600 jobs. Professional & Business Services and Educational & Health Services had the overall largest gain. Arts, Entertainment & Recreation employment declined for the second month in a row – the only sector that had a decline in July.
July 2015 parts.png
July 2015 pie.png
For background on the methodology, see our working paper Employment in tangible-producing and intangible producing
industries
.

Going beyond technical to other intangible skills

One of the key points of the intangible capital framework is that it stresses the importance of multiple forms of capital. Technical assets and skills, embedded in both human capital and structural capital, are only part of the equation. This example (from a Forbes article “That ‘Useless’ Liberal Arts Degree Has Become Tech’s Hottest Ticket”) about OpenTable’s experience illustrates the point:

Suddenly OpenTable needed salespeople. Years of selling helped OpenTable slip its software into more than 10,000 restaurants by 2008. That was a fragile triumph, however. Open-Table’s engineers kept upgrading the company’s seating systems and data analytics, only to discover that restaurateurs weren’t paying attention. That created a greater risk of customer churn. If OpenTable wanted strong, lasting connections with restaurant managers and owners, it needed a second team of frontline relationship-builders.
So OpenTable executives began hunting for people who had waited on tables, tended bar or managed restaurants earlier in their careers. The company was moving beyond its beginnings as an automation tool. The new priority, as sales chief Mike Dodson explains, was to find or train evangelists who could “show how tech can enrich the dining experience.”

In other words, relational capital, tacit knowledge and what is often called “social intelligence” are just as important to success in the Intangible Economy.

Agriculture as an information intensive industry

A number of years ago I wrote a report for the Economic Development Administration on Knowledge Management as an Economic Development Strategy. In the report I made the point that rural areas and agriculture are part of the information/intangibles economy by noting the following example of using local knowledge as an economic advantage:

For example, AgriImaGIS is an agricultural imaging business run by a former farmer. His knowledge of how to translate the information from satellite images into information for farmers on vegetation density, crop quality and the specific needs for fertilizer and pesticides was gained through 20 years of farming. (Cited from a story in the San Jose Mercury News

Earlier this week the New York Times ran a piece on “The Internet of Things and the Future of Farming”. The story describes a recent conference presentation by Lance Donny of OnFarm Systems:

In his presentation, Mr. Donny placed the progression of farming in three stages. The first, preindustrial agriculture, dating from before Christ to about 1920, consisted of labor-intensive, essentially subsistence farming on small farms, which took two acres to feed one person. In the second stage, industrial agriculture, from 1920 to about 2010, tractors and combine harvesters, chemical fertilizers and seed science opened the way to large commercial farms. One result has been big gains in productivity, with one acre feeding five people.
The third stage, which Mr. Donny calls Ag 3.0, is just getting underway and involves exploiting data from many sources — sensors on farm equipment and plants, satellite images and weather tracking. In the near future, the use of water and fertilizer will be measured and monitored in detail, sometimes on a plant-by-plant basis.
Mr. Donny, who was raised on a family farm in Fresno, Calif., that grew table grapes and raisin grapes, said the data-rich approach to decision making represented a sharp break with tradition. “It’s a totally different world than walking out on the farmland, kicking the dirt and making a decision based on intuition,” he said.
The benefits should be higher productivity and more efficient use of land, water and fertilizer. But it will also, Mr. Donny said, help satisfy the rising demand for transparency in farming. Consumers, he noted, increasingly want to know where their food came from, how much water and chemicals were used, and when and how it was harvested. “Data is the only way that can be done,” Mr. Donny said.

With cloud computing, intelligent software and cheap sensors, the Internet of Things will only increase the data available. Combined with autonomous vehicles and robotics, agriculture will become even more information intensive. (See also my earlier posting.)
PS – the company I cited in the report is still doing agricultural imaging:

Agri ImaGIS, working with a German satellite imagery company, uses a software mapping system to make satellite images more useful to farmers. The ag producers use the information to improve their farming operation, particularly in fine-tuning the amount of seed and chemical they apply.

June trade in intangibles

This morning’s trade data from BEA shows a widening deficit as imports jumped by $2.8 billion and exports remaining basically flat (dropped by $100 million). The deficit grew from $40.9 billion in May to $43.8 billion in June. The size of the deficit was pretty much in line with economists’ estimates of $43 billion. A major contributor was a big jump in petroleum goods imports, reversing a downward trend for most of this year. [Note that last week the Census Bureau’s advance estimate showed the goods trade deficit jumping from $59.7 billion in May to $62.3 billion in June as export were down and imports up (see earlier posting). Today’s data shows a goods trade data of $63.5 billion.]
However, once again exports of pure intangibles continued to grow. While imports also rose slightly, the intangibles surplus increased to $15.7 billion in June. Business services exports grew faster than imports. However, continuing a worrisome trend, net revenues from the use of intellectual property dropped slightly as revenues from foreign sources (exports) were up but charges for the use of intellectual property paid out to foreign sources (imports) increased more. The surpluses in maintenance & repair services and in financial services grew with both exports and imports up. The deficit in insurance services improved slightly, while the very small surplus in telecommunications services declined. (See detailed charts below.)
Our Advanced Technology deficit rose slightly from $7.2 billion in May to $8.8 billion in May. Imports increased in all Advanced Technology industries except Flexible Manufacturing. Exports were generally up in almost every industry but not enough to counter the import surge. The biggest change in the deficit came from an increase in Information and Communications Technology (ICT) imports.
Advanced Technology goods also represent trade in intangibles. These goods are competitive because their value is based on knowledge and other intangibles. While not a perfect measure, Advanced Technology goods serve as an approximation of our trade in embedded intangibles. Adding the pure and embedded intangibles shows an overall surplus of $6.9 billion in compared with $8.3 billion in May.
Intangibles trade-June15.png
Intangibles and goods-June15.png
Oil goods intangibles-June15.png
Intangibles trade parts-June15.png
Business services-June15.png
IP-June15.png
Financial services-June15.png
Maintenance and repair services-June15.png
Insurance services-June15.png
ICT-June15.png

Note: I am now reporting the trade data using the new BEA classifications for services trade, which breaks services into more categories. In the past, the intangible trade data was the sum of Royalties and License Fees and Other Private Services. Under the new classification system, intangibles trade data is the sum of the following items: maintenance and repair services n.i.e. (not included elsewhere); insurance services; financial services; charges for the use of intellectual property n.i.e.; telecommunications, computer, and information services; other business services.


Charges for the use of intellectual property n.i.e. is simply a renaming of Royalties and License Fees. This includes transactions with foreign residents involving intangible assets and proprietary rights, such as the use of patents, techniques, processes, formulas, designs, know-how, trademarks, copyrights, franchises, and manufacturing rights.


Maintenance and repair services n.i.e., financial services, and insurance services, were previously included in Other Private Services. Telecommunications, computer, and information services is a combination of those two items (telecommunications and computer & information services) that were also previously included in Other Private Services. Three categories previously in Other Private Services — education-related and health-related travel and the expenditures on goods and services by border, seasonal, and other short-term workers — were removed and reclassified to travel. The new category of other business services is a continuation of the older category Other Private Services with those components removed.


Thus, other business services includes categories such as advertising services; research, development, and testing services; management, consulting, and public relations services; legal services; construction, engineering, architectural, and mining services; and industrial engineering services. It also includes personal, cultural, and recreational services which includes fees related to the production of motion pictures, radio and television programs, and musical recordings; payments or receipts for renting audiovisual and related products, downloaded recordings and manuscripts; telemedicine; online education; and receipts or payments for cultural, sporting, and performing arts activities.


For more information on the changes, see the March 2014 Survey of Current Business article, “The Comprehensive Restructuring of the International Economic Accounts: Changes in Definitions, Classifications, and Presentations.”

New data series on goods trade balance

Starting today, the Census Bureau is publishing a new advanced trade data series. This advanced data covers trade in goods only. (See Census Bureau notice.) Since this excludes services and intangibles trade, I will not be covering this data carefully
The next release of the full set of trade data is next week. I will publish my analysis of intangibles trade then.
According to Census, the goods trade deficit jumped from $59.7 billion in May to $62.3 billion in June. Export were down and imports up.

2Q 2015 GDP – advanced data

GDP data released this morning from BEA shows the US economy growing at an annual rate of 2.3%. Economist had predicted an annual GDP growth rate of 2.5%. This rate for the second quarter of 2015 is a large increase over the anemic 0.6% for the first quarter. And that revised figure for the first quarter was better that the previous data showing a contraction in the economy.
Investment in Intellectually Property Products (IPP) slowed down slightly to 4.9% from 7.4% in the 1st quarter of 2015 and 7.4% in the 4th quarter of 2014. [This is a revision downward for earlier quarters from previously published data.] R&D grew by 5.2% compared to 6.9% in 1Q and 8.8% in the 4Q. Investment in software grew by 7.8% compared to 9.1% in 1Q and 5.6% in 4Q. On the other hand, investments in entertainment, literary, and artistic original dropped by 2.4% while growing by 2.2% in 1Q and 4.9% in 4Q.
Today’s BEA release also contains data revisions going back to 1969. This includes earlier data on Intellectually Property Products. More analysis later of both the revisions and what the data tells us about previous decades.
IPP parts 2Q15 - 1st.png