Peter Josty is Executive Director of The Centre for Innovation Studies based in Calgary.

Most statistical agencies – including Statistics Canada – distinguish between the manufacturing sector and the services sector in their data collection and reporting. With all the changes that have taken place in the economy in recent years, is this distinction still meaningful and useful?

The traditional manufacturing company buys its raw materials, conducts R&D, manufactures products in its own factory and then sells the product to customers.

But manufacturing isn’t what it used to be. Some manufacturers don’t make anything themselves (Apple, for example); some manufacturers sell services as well as products (Rolls-Royce jet engines); and some products are sold as a subscription (software as a service).

This is very different from Henry Ford’s Rouge River plant where he created a self-sufficient industrial complex where raw materials like iron ore, coal and limestone were processed and transformed into finished vehicles.

However, Diane Coyle, a distinguished British economist, writes in a recent book, The Measure of Progress: Counting What Really Matters:

“The traditional distinction between manufacturing and services is not only meaningless but actively unhelpful for understanding the economy.”

What has changed?

1. “Factoryless” goods production (FGP)

Firms in this category keep the high value-added work in-house and contract out the lower value-added stages of production, sometimes overseas. Because of this they tend to have higher productivity than other firms in their sector.

The pharmaceutical industry is one of the biggest users of FGP, where the pharmaceutical company owns the intellectual property and contracts out the manufacturing to others.

Most pharmaceutical companies use contract manufacturing to some extent. There are a number of rationales for doing this – to avoid risky capital investment, have flexibility when future demand is uncertain, and saving costs, as the contract manufacturer can work for several producers and achieve economies of scale.

The auto sector is another big user of FGP. Magna International is a major contract manufacturer, making cars for Mercedes-Benz, Jaguar and BMW.

Other examples include Apple, which contracted most iPhone production to Taiwan-based Foxconn, and semiconductor makers, which contracted with the Taiwan Semiconductor Manufacturing Company to manufacture advanced chips.

2. “The subscription economy”

In this category, firms own the whole manufacturing chain but add value-added services to the final product, selling a bundle of combined products plus services to the end customer.

An example is Rolls-Royce, which manufactures jet engines for airplanes. Instead of the customer buying the engine and arranging for maintenance and support, Rolls-Royce sells a long-term contract that has a fixed fee for a certain number of flight hours. They monitor the engines 365 days a year and carry out all the maintenance and support.

This is known as “power by the hour.” This is a big cost saving for the airline, as they avoid the cost of spare parts inventory, repair facilities, technicians and engine liability insurance. Because of economies of scale, Rolls Royce can provide these services at a lower cost than individual airlines.

Another example is lighting. Phillips Lighting sells “light as a service,” so instead of a customer buying light bulbs and changing them when they expire, Phillips provides a service to maintain constant lighting.

This model has proliferated a lot in recent years. “Software as a service” is now becoming the norm, where instead of buying a computer program you buy a subscription and the supplier updates the program periodically.

Cloud computing is another example of the subscription economy. Instead of a company buying its own servers, it contracts that service to another company that provides that service for many others, which results in lower costs all round.

Because of the enormous costs and economies of scale this market has become very concentrated, with Amazon Web Services becoming the dominant supplier.

Good data is not available on this approach, but Coyle estimated that about half of all manufacturers in OECD countries use the subscription economy model to some extent.

As she wrote: “In the product economy it was all about things. Acquiring new customers, shipping commodities, billing for on-time transactions. But in the new era it’s all about relationships. More and more customers are becoming subscribers because it meets their needs better.”

There are some wrinkles with the subscription model. For example, John Deere and General Motors sell products, but retain all the software IP in these products and refuse to share it. So, a customer is not able to repair their purchase. This has led to a “right to repair” movement.

This approach is also used for some medical equipment, ink for printers, and many other areas to limit competition and raise prices for customers. This has been colourfully described as enshittification – the progressive worsening of things that used to work well because of the market power of the supplier.

Of course, the subscription business model has been around for a long time – think leasing and newspapers – but it has become much more widespread.

Does this matter?

It has been estimated that the rise of factoryless goods production and the subscription economy has undercounted the manufacturing sector in OECD economies and reduced estimates of GDP.

We lack good data on these phenomena and that hinders economic policymakers.

Conclusion

The distinction between manufacturing and services has become blurred by the rise of factoryless goods production and the subscription economy to the point where it is actively unhelpful. Both of these trends have been enabled by widespread use of information technology.

There are many beneficial effects of these changes but also some malign ones, such as enshittification and the creation of significant market concentration in some markets (e.g. with Amazon Web Services).

We lack good information on these structural changes, and better data needs to be developed to help with efforts to increase productivity and growth.

R$