Disclaimer: I am now retired, and am therefore no longer an expert on anything. This blog post presents only my opinions, and anything in it should not be relied on.
In my view, Haskel/Westlake’s “Capitalism Without Capital” is not so much an argument that the increasing importance of “intangible assets” constitutes a new and different “intangible economy”, as strong evidence that the ever-increasing impact of software means that a fundamental idea of economics – that everything can be modeled as a mass of single-product manufacturers and industries – is farther and farther from the real world. As a result, I would argue, measures of the economy and our well-being based on those assumptions are increasingly distorted. And we need to do more than tweak our accounting to reflect this “brave new world”.
First, my own brief “summary” of what Haskel/Westlake say. They start by asserting that present-day accounting does not count intangible company investments like software during development, innovation property such as patents and R&D, and “economic competencies” such as training, branding, and business-process research. In the case of software development, for example, instead of inventory that is capitalized and whose value is represented at cost until sold, we typically have expenses but no capitalized value right up until the software is released and sold. A software company, therefore, with its continual development cycle, appears to have zero return on investment on a lot of its product portfolio.
Haskel/Westlake go on to argue that a lot of newer companies are more and more like software companies, in that they predominantly depend on these “intangible assets.” The new breed of company, they say, has key new features:
1. “sunk costs” – that is, you can’t resell development, research, or branding to get at its monetary value to you.
2. “spillovers” – it is exceptionally easy for others to use your development-process insights and research.
3. “scalability” – it requires relatively little effort and cost to scale usage of these intangible assets from a thousand to a million to a billion end users.
4. “synergies” – research in various areas, software infrastructure, and business-process skills complement each other, so that the whole is more than the sum of the value-added of the parts.
5. “uncertainty” – compared to, say, a steel manufacturing firm, the software company has far more potential upside from its investments, and often far more potential downside.
6. “contestedness” – such a company faces much greater competition for control of its assets, particularly since they are so easy to use by others.
Finally, Haskel/Westlake say that, given their assumption that “intangible companies” make up a significant and growing part of the global economy, they already have significant impacts on that economy in particular areas:
· “Secular stagnation” over the last decade is partially ascribed to the increasing undervaluing of these companies’ “intangible assets”.
· Intangible companies increase income inequality because they function best with physical communication by specialized managers in cities.
· Intangible companies are under-funded, because banks are not well suited to investing without physical capital to repossess and resell. Haskel/Westlake suggests that greater use of equity rather than loans is required, and may be gotten from institutional investors and by funding collaborating universities.
· Avoiding too much failure will require new business practices as well as new government encouragement, e.g., via better support for in-business control of key intangible assets (clear intangible-asset ownership rules) or supporting the new methods of financing the “intangible companies.”
It’s the Software, Sirs
Let’s look at it a different way. Today’s theories of economics grew out of a time (1750-1850) when large-scale manufacturing was on the rise, and its microeconomics reflects that, as does the fact that data on economic performance (e.g., income) comes from surveys of businesses, which is then “adjusted” to try to include non-business data (trade and reconciling with personal income reports). From 1750-about 1960, manufacturing continued to increase as a percentage of overall economic activity and employment, at the expense of farming. From 1960 or so, “services” (ranging from hospitals to concierges) began to carve into that dominance, but all those services, in terms of jobs, could still be cast in the mold of “corporation that is mostly workers producing/dealing with customers, plus physical infrastructure/capital”.
Now consider today’s typical large software-driven company. Gone is the distinction between line and staff. Manufacturing has shrunk dramatically as a share of economic activity, both within the corporation and overall. Bricks and mortar is shrinking. Jobs are much more things like developers (development Is not manufacturing nor engineering but applied math), marketers/branders, data scientists (in my mind, a kind of developer), help desk, Internet presence support. The increased popularity of “business agility” goes along with shorter careers at a particular company, outsourcing, intra-business “services” that are primarily software (“platform as a service”, Salesforce.com). Success is defined as control over an Internet/smartphone software-related bottleneck like goods-ordering (Amazon), advertising (Google), or “apps” (Apple).
Now consider what people are buying from these software-driven firms. I would argue that it differs in two fundamental ways from “manufacturing” and old-style “services”:
1. What is bought is more abstract, and therefore applicable to a much wider range of products. You don’t just buy a restaurant meal; you buy a restaurant-finding app. You don’t just browse for books; you browse across media. What you are selling is not a widget or a sweater, as in economics textbooks, but information or an app.
2. You can divide up the purposes of buying into (a) Do (to get something done); (b) Socialize/Communicate, as in Facebook and Pinterest; and (c) Learn/Create, as in video gaming and blog monetization. The last two of these are really unlike the old manufacturing/services model, and their share of business output has already increased to a significant level. Of course, most if not all software-driven companies derive their revenues from a mix of all three.
The result of all this, in economic terms, is complexity, superficially masked by the increased efficiency. Complexity for the customer, who narrows his or her gaze to fewer companies after a while. Complexity for the business, whose path to success is no longer as clear as cutting costs amid stable strategies – so the company typically goes on cutting costs and hiring and firing faster or outsourcing more in default of an alternative. Complexity for the regulator, whose ability to predict and control what is going on is undercut by such fast-arriving devices as “shadow banking”, information monopolies, and patent trolling.
In other words, what I am arguing for is a possible rethinking of macroeconomics in terms of different microeconomic foundations, not the ones of behavioral economics, necessarily, but rather starting from the viewpoint “what is really going on inside the typical software-driven corporation” and then asking how such a changed internal world will reflect back to the overall economy, how macroeconomic data can capture what is going on, and how one can use the new data to regulate and anticipate future problems better.
John LeCarre once said that the key problem post-Cold-War was how to handle the “wrecking infant” – here he was referencing an amoral businessman creating Third-World havoc, although you can translate that to the situation in the US right now. The software-driven business, in terms of awareness of what to do, is a bit of a wrecking infant. If it isn’t helped to grow up, the fault will not lie solely in our inability to anticipate the distorting rise of intangible assets, its side-effect, but also in our failure to deal adequately with its abstraction, new forms of organization and revenue, and complexity.