“New data,” writes Christopher Mims of The Wall Street Journal, “suggests that the secret of the success of the Amazons, Googles and Facebooks of the world — not to mention the Walmarts, CVSes and UPSes before them — is how much they invest in their own technology.” This new analysis from James Bessen of Boston University shows a big productivity divergence between the top 5 percent of manufacturing and service firms and everyone else. And it tends to reinforce the thesis Michael Mandel and I have offered that the “information gap” is key to the productivity slowdown.
The policy conclusion Mims and Bessen draw, however, is that today’s information technology (IT) dominance will translate to long-term economic concentration. As Mims concludes, “A select few monopolize the gains, while many fall increasingly behind. Might it eventually be the case that the biggest firms aren’t just dominant, but all-encompassing?”
It’s not a crazy idea, but I think it suffers from a static view of the economy. Yet as these firms have shown in a relatively short period of time, the economy is constantly changing. At least that’s the hope.
But back to the IT analysis. Mims says:
There are different kinds of IT spending. For the first few decades of the PC revolution, most companies would buy off-the-shelf hardware and software. Then, with the advent of the cloud, they switched to services offered by the likes of Amazon, Google and Microsoft. Like the difference between a tailored suit and a bespoke one, these systems can be customized, but they aren’t custom.
IT spending that goes into hiring developers and creating software owned and used exclusively by a firm is the key competitive advantage. It’s different from our standard understanding of [research and development] R&D in that this software is used solely by the company, and isn’t part of products developed for its customers.
The economic data have indeed shown a recent spike in spending on non-capital technology goods and services by non-tech companies. For example, info-tech spending in the energy industry has recently surged, and so has its productivity.
That productivity gap correlates with the increase in spending on proprietary IT, says Mr. Bessen. In 1985, firms spent on average 7% of their net investment (which includes software, new buildings, R&D and the like) on proprietary IT, according to data from the Bureau of Economic Analysis. In 2016, about 24% of U.S. firms’ net investment went into proprietary IT. That’s nearly $250 billion in a single year, and almost matches their outlay for R&D and capital expenditures.
This points toward another interesting phenomenon — the so-called end of R&D. Amazon, for example, long ago stopped thinking about R&D as a separate operational and financial category. Instead, it considers nearly everything it does as some combination of investment in research, development, and product that doesn’t fit into the old accounting buckets. It rolls engineering compensation and hardware and software spending into a larger category labeled “technology and content.” Partly in response to this changing landscape, the Bureau of Economic Analysis just published a major revision of economic data and updated its methodology to measure internal software and information investments.
So does the information gap solidify the current tech giants at the top? Or does the gap show something like the opposite? Might it suggest a gigantic opportunity for the rest of the economy to more fully exploit these cheap information tools to remake their businesses and for entrepreneurs to start new ones?
Just as Amazon and Google used microchips, hard drives, and internet connectivity to build entrepreneurial businesses that delivered old products in new ways and entirely new products, the next wave of entrepreneurs will exploit new technologies such as artificial intelligence, CRISPR and the exploding world of bio-information, blockchain, crypto assets, the Internet of Things, robotics, computer vision, and 5G wireless to build the giant companies of tomorrow. This optimistic outlook is not guaranteed, however. If we enact policies that assume a calcified world of top-heavy technological concentration, we can expect a counterproductive outcome. Assuming that monopolies exist is often a self-fulfilling prophecy.