Vox’s economics correspondent Timothy Lee is worried about Uber. Not for the usual reasons that have been debated to death, like pay and regulation. His concern is more novel, and very interesting.
What Uber is (or might be) doing
Uber is raising and investing lots of money, Lee points out. Only this week, it raised a massive $3.5 billion from the Saudis, and made it look casual. Normally this should be good news for the economy. After all, everyone’s worried about low investment – it’s the hallmark of so-called secular stagnation, the phenomenon that economic galacticos like Larry Summers argue is holding back productivity. All other things being equal, we’d like to see firms investing more. Especially smart firms like Uber who, we hope, will make canny investments.
But from the point of view of the economy as a whole, not all profitable investments are equal. Some investments benefit not just the business who makes them, but also other businesses – even competitors. Innovation investments are often like this. Apple invented the iPhone, but for all their patents and proprietary systems, their investment also benefited Samsung, Google and thousands of other firms. As economists would say, it was an investment with large positive spillovers. The gap between them is one of the reasons governments around the world subsidise businesses to do R&D.
Uber’s recent investments, Lee argues, are not much like this. They’re mainly about giving big discounts to users and payouts to new drivers to enlarge the Uber network in places China.
Under some circumstances, you could imagine that these network-building investments do have positive spillovers. For example, if Uber lobbies the Chinese government to allow certain forms of ridesharing, or invests in marketing to convince people that ridesharing is safe and classy, that would benefit not just Uber but also its ridesharing competitors by growing the overall market.
But according to Lee, Uber’s spending is mostly not about this – it’s more about winning tooth-and-nail battles for dominance with competitors like Lyft and Didi Chuxing.
These could still be very good investments for Uber’s shareholders. Ride-sharing is probably one of those winner-takes-all businesses where building the biggest network yields huge profits and the runners up can go whistle. As economists would say, there are “network effects” at work. Investing $3.5 billion of Saudi money to win these network wars may pay off handsomely. But it’s very possible that it won’t generate large spillovers in the way that Tesla spending $700m on battery R&D would.
Uber’s network: an intangible asset?
It’s worth reflecting on how we think about what Uber is doing with its vouchers and driver subsidies. As Lee says, this spending is clearly an investment, in that Uber is spending money today to create something that, they hope, will let them reap rewards in the longer term.
It’s also clear that this investment is intangible: it consists of the relationship and the contracts between drivers and Uber, not of physical stuff like lathes or servers or office buildings. But at the same time, it’s clear that Uber is not creating a public good that anyone can freely use like Pythagoras’s Theorem or oral rehydration therapy – the Uber network is just that: it’s Uber’s network.
In the early 2000s, economists were trying to get to grips with what was then called, without cynicism, the “knowledge economy” or the “new economy”. Carol Corrado, Charles Hulten and Dan Sichel set out a framework for how we might measure what was going on in a world ever more dependent on ideas.
They came up with a methodology for measuring intangible investment – spending that firms did to bring them long-term benefit, but that related to ideas and relationships rather than physical things. It gradually became clear that intangible investment was increasing rapidly; indeed, by the late 2000s countries like the US and the UK were spending more on intangible investment than on tangibles.
This included things like software and databases (they called this “computerized information”); R&D, product design and artistic originals (“innovative property”) and business-related investments like organizational development, marketing and firm-specific training (“economic competencies”).
The last category mainly referred to things going on within the firm: the original paper refers to “time spent on improving the effectiveness of business organisations”. But since the paper was published some interesting changes happened in the economy as whole. One is the sharing economy, in which companies like Uber and AirBnB use the Internet to build networks of not-employees on whom their business depends. Another is the smartphone revolution, one part of which involves Apple and Google cultivating ecosystems of app developers.
Of course, relations outside the firm have always been valuable, supply chains being one big example. But in recent years we’ve seen the emergence of large, high-profile companies whose main asset is the result of what Corrado, Hulten and Sichel would call organizational development investment, but outside the walls of the firm itself.
What is happening to spillovers?
This brings us back to Lee’s original point. Lee is arguing that Uber is spending its newly raised money on intangible investments that generate very few spillovers or none at all. As a result, productivity in the economy as a whole will increase less than if they were spending the same money on, say, an amazing R&D project.
Some intangible investment seems to have high spillovers. People are so convinced of the spillovers of R&D that they are willing to pay taxes to subsidise companies to do more of it. But it’s not just R&D: in the example of the iPhone we discussed earlier, Apple’s design, software and marketing investments also helped make prepare the market for smartphones in general, making the world safe for Android and the Samsung Galaxy.
Lee’s Uber example, on the other hand, suggests that some intangibles have very few spillovers. (You can imagine other intangible investments about which this would also be true. If I program a HFT algorithm to be faster than your HFT algorithm, there might not be any spillovers. A marketing campaign in a low-growth market might be mostly about winning market share – which could also imply low spillovers. Equally, it’s worth noting that other aspects of Uber’s business create positive spillovers, such as reductions in drunk-driving accidents.)
This raises an interesting question. What if the balance between high-spillover intangible investment and low-spillover intangible investment was also shifting? What if there were more Uber-style tournaments and fewer Tesla battery moonshots?
Intuitively, you’d expect productivity growth to slow, even if overall levels of intangibles were increasing. It’s not clear whether this is happening or not, but in an age of productivity worries, it would be interesting to investigate it more – and if it is happening, to consider how we can change it.