A version of this post originally appeared on Forbes.com

Can digitizing the food manufacturing industry help boost living standards? The short answer is yes, if we link food manufacturing into the Internet of Goods.

We’re used to thinking of food as cheap and getting cheaper. In 1947, spending on food—both in and out of the home—accounted for 27% of non-health personal spending. By 2000, the food budget share, omitting healthcare, had dropped to 14%.

This 50-year decline in the food budget share fueled American prosperity. With much less of their budgets going to food, middle-class households could afford to spend more on housing, cars, vacations, and all the other aspects of a good life.

But as Figure 1 shows, around 2000, something changed. The decline in the food budget share stopped. Indeed, household spending on food has inched up to close to 15% of non-health personal spending by 2017.

To put it another way, if the past trends had continued,  the food share of non-health spending would be only 10%.  Americans would have almost $500 billion more to spend in other areas.

What happened to the food industry? Groups such as the American Antitrust Institute point to consolidation in industries such as meat processing, which potentially has increased the market power of major players and their ability to raise prices.

Another factor boosting food costs may be greater attention to safety.  In particular the Food Safety Modernization Act (FSMA) was signed into law by President Barack Obama on January 4, 2011. This legislation gives the FDA a new mandate to regulate food production and processing. Indeed, the necessary rules are still being implemented–for example, the FDA is currently asking for comments on a proposed guidance for “Standards for the Growing, Harvesting, Packing, and Holding of Produce for Human Consumption.” Obviously  an important step given the current lettuce issue!

But perhaps most important, the food manufacturing industry has been in a deep and profound productivity slump in recent years. Measured by the Bureau of Labor Statistics (BLS), output per hour in food manufacturing has dropped by 8% since its peak in 2005 (Figure 2).

In response, the food manufacturing industry has been embracing digitization, but it’s a slow process. According to data from the BLS, the entire food manufacturing industry employed less than 1000 software developers and programmers as of May 2017, compared to 25,000 engineers and scientists.

Digitization will have a significant impact in several different areas of food manufacturing. First, it will become much easier to consistently track food from “farm to fork.”  As a result, food recalls will become easier and cheaper. (According to one count, there were 456 food recalls in 2017).

Second, digitization of the production process will help boost productivity and lower costs. This includes product development. For example, FlavorWiki is a startup that uses data analytics to quantify consumer taste perceptions, and potentially help companies develop new products.

Finally, and perhaps the most important, digitization allows the development of local production models for food, requiring food to be shipped much shorter distances. In Japan, for example, the world’s largest automated leaf-vegetable factory has just opened in a suburb of Kyoto. In the U.S., companies such as Iron Ox are developed autonomous and hydroponic production models.

Such vertical farms might be tied directly into ecommerce networks to handle local delivery direct to consumers, thus cutting out several layers of the distribution chain. The result would be lower prices, higher quality, and less pollution from shipping, These are some of the benefits of the Internet of Goods.

Indeed, digitization will enable the rethinking of the entire food production, manufacturing, and distribution chain, to the benefits of consumers. With any luck, Americans will once again find their food budget shares falling and their standard of living rising.

Summary: Based on new “digital economy” data from a recent BEA working paper, we calculate that the labor share of the digital sector has risen since 2007, while gross margin of the digital sector has fallen over the same period. This result is consistent with strong competition in the digital product and labor markets (see the research note here).

In March 2018, the Bureau of Economic Analysis (BEA) released a working paper called “Defining and Measuring the Digital Economy.”[1] The working paper presented BEA’s initial work “to lay the foundation for a digital economy satellite account.”

The BEA authors focus on outlining their definition of the digital economy, and calculating its real growth and share of GDP. However, their data allows us to calculate two other policy-relevant measures of the digital economy: Labor share and gross margin.

 Labor share is a measure of how much of the income of an industry is going to workers. For the purposes of this paper, we define the labor share as compensation (COMP) divided by value-added (VA), expressed as a percentage. [2]

Gross margin is a measure of the profitability of an industry per unit of sales. In the business literature, gross margin is a company’s total sales revenue minus its cost of goods sold, divided by total sales revenue, expressed as a percentage.[3]

For our purposes, we define gross margin as an industry’s total gross output (GO), minus the cost of intermediate inputs (II) and labor compensation (COMP), divided by total gross output, expressed as a percentage.[4]

Based on this definition, labor share in the private sector has trended down since at least 1990 (Table 1). Similarly, private sector gross margin have trended up since at least 1990. Since 2007, private sector labor share has fallen by 0.8 percentage points, and private sector gross margin has risen by 1.9 percentage points.[5]

Table 1: Private Sector: Falling Labor Share, Rising Gross Margin
	1990	2007	2016
Labor Share	52.2%	50.6%	49.8%
Gross Margin	26.0%	26.7%	28.6%
Data: BEA (as of April 2018)

 

Table 2: Digital Sector: Rising Labor Share, Falling Gross Margin
		2007	2016
Labor Share		53.4%	55.4%
Gross Margin		28.4%	27.2%
Private sector industries only. Data: BEA working paper

The digital economy data from the BEA allows us to calculate the labor share and gross margin for the digital sector of the economy (Table 2). We see that labor share for private industries in the digital sector rose by 2 percentage points in the post-2007 “tech boom” period. Gross margin fell by 1.2 percentage points.

Figures 1 and 2  show the change in the labor share over time. Please note that this data was released prior to the July 2018 benchmark revision.

These results suggest that benefits of productivity growth in the digital sector since 2007 are being shared with workers and customers. This is consistent with strong competition in the digital product and labor markets. By contrast, companies in the broader private sector are benefitting from lower labor share and higher gross margin, which suggest that market power is rising outside of the digital sector.

The PDF version of this research note are found here. The results in this note are drawn from a forthcoming PPI policy paper, “Taking Competition Policy Seriously.” 

 

Definition of Measures

Labor share = COMP/VA

Gross margin = (VA-COMP)/GO = (GO- II-COMP)/GO

VA = value-added

COMP = labor compensation

GO = Gross output

II= Intermediate inputs

 

[1] Kevin Barefoot, Dave Curtis, William Jolliff, Jessica R. Nicholson, Robert Omohundro. “Defining and Measuring the Digital Economy,” March 2018. https://www.bea.gov/digital-economy/_pdf/defining-and-measuring-the-digital-economy.pdf

[2] Several alternative measures of the labor share all have the same general trend.

[3] https://www.investopedia.com/terms/g/grossmargin.asp#ixzz5No688Apd

[4] The numerator includes profit-type income, such as profits, rents, and interest. It also includes taxes on production and imports that are chargeable to business expenses, such as state and local sales and property taxes, and a hodgepodge of state, local, and federal excise taxes.

[5] All data in this note is prior to the July 2018 benchmark revision. We focus only on private industries.

The FCC, under Chairman Ajit Pai,  voted today to start the process of eliminating Title II rules on ISPs. We applaud his move. As we have said before, the Internet was thriving under the light-touch regulatory regime that preceded Title II. Indeed, government data shows that the telecom industry was one of the top contributors to US productivity growth from 2000 to 2014, before Title II was put in place. *

Our belief is that the economy could be entering into a renewed period of productivity growth, propelled by the application of digital technology to the physical industries (see The Coming Productivity Boom). That transition would have been much more difficult under the antiquated regulatory structure that comes with Title II.

But as we have also said before, smart regulations are essential for a well-functioning economy. There’s no doubt in our mind that open internet rules are needed to govern the Internet. We think now is the right time for Congress to codify the open internet principles into law–that’s the best way to get a consistent regulatory structure that will produce faster growth for us all.

 

*Contribution to multifactor productivity growth, as calculated from the BEA-BLS Integrated Production Accounts.

 

 

 

Here are some bullet points on the economics of the election:

1. America is divided between the Digital Nation and the Physical Nation. The Digital Nation includes tech, entertainment, publishing, telecom, finance, and professional services such as management consulting, accounting, computer programming, design. The Physical Nation includes manufacturing, mining, construction, retail, transportation, health care, and the rest of the economy The Digital Nation makes up about 25% of private sector employment, the Physical Nation 75% (we first laid out this division of the economy in a March 2016 report).
2. While there are many factors going into Trump’s election, on the economic side, there was one reality:  The members of the Physical Nation finally got tired of suffering  while the Digital Nation soared. And since the Physical Nation outnumbers the Digital Nation 3-1, that’s the election.
3. For the past fifteen years, the Digital Nation enjoyed strong productivity growth, stable prices, high investment in IT, rising employment, and higher (and rising) incomes. By contrast, the Physical Nation has suffered from weak productivity growth, rising prices, weak investment in IT, weak employment growth (outside of healthcare), and lower (and barely rising) incomes.

Digital Nation vs Physical Nation
	Digital Nation	Physical Nation
Productivity growth rate (2000-2015)	2.7%	0.7%
Real compensation per worker,  growth rate (2000-2015)	1.3%	0.8%
Employment growth rate (2000-2015)	1.3%	1.4% 0.1% (without healthcare)
Share of private sector employment (2015)	25%	75%
Share of private sector compensation (2015)	35%	65%
Share of IT investment (2015)	75%	25%
Annual price change	0.8%	2.4%
Data: BEA, BLS, author calculations The split between the digital and the physical sector was first described in Mandel (2016). Numbers may differ slightly from earlier calculations.

 

4. The Digital Nation is concentrated in blue states. States that voted for Clinton in this election averaged 35% digital, while states that voted for Trump are 23% digital on average. Here are the top states, measured by share of private sector GDP coming from the digital sector.

 

Top Digital States
	Share of private economy that is digital
DC	49.9%
Delaware	47.8%
New York	43.8%
Massachusetts	37.7%
Oregon	37.4%
Connecticut	34.3%
Virginia	33.5%
California	33.5%
Colorado	32.5%
Rhode Island	31.5%
Maryland	31.1%
Georgia	30.8%
NH	30.4%
Illinois	30.1%
New Jersey	29.8%
Minnesota	29.8%
Washington	29.4%
Missouri	28.0%
North Carolina	27.7%
Utah	27.6%
Pennsylvania	27.2%
Arizona	26.8%
Florida	26.5%
South Dakota	26.3%
Ohio	25.8%
Nebraska	23.5%
Kansas	23.3%
Michigan	23.3%
Wisconsin	23.2%
Data: BEA, author calculations

 

 

Next: How trade and productivity growth have affected the Physical Nation