The single most important development in the digital economy over the last decade has been the architectural transformation from on-premise servers and software to “the cloud,” including computing and software as a service. Nothing has been more significant to reducing the costs of start-up innovation, enabling rapid scaling up and down of computing capacity in sync with demand and making progress on cybersecurity. This cloud-enabled digital infrastructure will be essential to making next generation machine learning applications broadly usable across the economy — which, in turn, is our best bet to generate the productivity we need to escape the risk of stagflation as we emerge from the COVID-19 crisis.
Given these high stakes, getting the terms of cloud competition right is now a critical economic and technology policy objective.
Competition in the cloud isn’t an easy problem to solve. Cloud providers are fundamentally a scale-advantaged business, where the bigger you are, the better. There are also likely some advantages to a certain amount of vertical integration — packaging raw computing power with integrated software packages, data security, and other related products all run on on-demand, can offer pricing and performance advantages. So, there are natural tendencies toward consolidation and, at a minimum, oligopoly.
But this is precisely why we have to be extra vigilant about the business models and licensing practices of the major cloud service providers.
If these platforms are set up to enhance monopolistic tendencies, it can and will exacerbate limits to competition, which will in turn reduce the positive impact that the cloud revolution can have on the overall economy.
We need to lean decisively in the other direction, and positively incentivize business models and practices that enhance competition from top to bottom of the cloud stack.
The European Union has been pushing in precisely this direction, and large technology companies have taken notice. In May, Microsoft publicly responded in a blog post titled “Microsoft’s European Cloud Principles.” Microsoft announced plans to relax a set of licensing restrictions that had had the effect of raising prices for customers using Microsoft enterprise software licenses on competing cloud services — but only for smaller cloud providers, and only for those in Europe.
This isn’t nearly enough.
Although Microsoft’s response was a reasonable half-step in the right direction, it falls short in crucial ways. Fair competition is not a game where you get to choose who you compete against.In this case, to move toward a truly level playing field would require that all cloud services — regardless of size and geography — be freed from discriminatory pricing practices. This impacts not just the direct competitors, but the economy overall.
As long as Google Cloud and Amazon AWS customers are still disadvantaged by Microsoft’s enterprise software pricing scheme, most European customers will continue to have to absorb higher prices for some important enterprise applications if they choose not to ‘switch’ to a bundle that lives in Azure.
What could possibly be the justification for limiting the change in licensing restrictions to small European-based cloud providers? And why should these principles apply in Europe but not elsewhere?
In fact, Microsoft’s offer is heavier on the public relations angle than the actual business economics angle, and it harkens back in unfortunate ways to the ‘bad old days’ of anti-competitive bundling practices which we simply cannot afford to allow to creep back into the cloud environment.
From a purely political and public relations standpoint, offering what looks like relief (though remember: it is relief from what was arguably an anti-competitive posture to start) targeted at smaller businesses in Europe could be a savvy move to try to reduce the immediate pressure from regulators and legislators in Brussels. But it doesn’t do enough to advance what ought to be the goal of cloud competition principles and policy.
The unfortunate fact is that bundling of applications with platforms can be an attractive way for big technology players to protect themselves from competition and raise the costs of switching to the point where it becomes prohibitive for customers to really even consider it. It’s attractive for the would-be monopolist, but it’s bad for customers and it’s bad for innovation.
Regulatory authorities in both the U.S. and Europe moved against this bundling strategy when it came to operating systems and web browsers a long time ago. Currently both are moving against this strategy again when it comes to app stores and apps, search and businesses that show up in search results and even e-commerce platforms and in-house products. These are important developments in competition policy, but they are also aimed at markets that are somewhat mature where the damage has already been done. Do we really want to wait for the same kind of problem to further impede the development of competition and innovation in the cloud?
Principles are one thing; policy, pricing, and licensing decisions that really affect how markets function are another.
Now is the right time to press for a better and fuller solution to the cloud competition challenge, one that supports broad economic growth and innovation more than public relations.
Steve Weber works at the intersection of technology markets, intellectual property regimes, and international politics. He has published numerous books, including “The End of Arrogance: America in the Global Competition of Ideas” and “The Success of Open Source,” and serves as professor of the Graduate School, School of Information, UC Berkeley. He has worked with and received research funding from a number of technology firms, including Google and Microsoft.