Monday, November 17, 2008

It's the Infrastructure, Stupid

(or how I learned to stop worrying and love socialism)

As we prepare for the beginning of a new presidency and an ascendant Democratic party, my thoughts have turned recently to governance and what lessons we can draw from the Open Source phenomenon (I refuse to call it a movement).

Infrastructure wants to be free

In the world of open source software, it became quite clear that matters of computing infrastructure, particularly operating systems, were easy pickings for open source processes. The reason for this is that no one wants to pay a premium for items that are taken for granted as the cost of doing business. And with the success of multiple open source projects, you no longer have to pay a premium for software that does the basics. The result is that there were plenty of reasons for a critical mass of people to get involved in the creation of these infrastructure items, ie. Linux, Apache, MySQL, et al. Because of this, you no longer find software companies looking to create proprietary versions of the above and charge a premium.

The success of the open source projects makes that business model obsolete. Instead, what the smarter companies have done is use these infrastructure components as the basis for the pieces that they build – which they can then charge a premium for. As a result, innovation happens because these companies are no longer saddled with the cost of creating infrastructure and can, instead, focus on the innovative pieces they wish to create. By bringing down the cost of innovation, it means that are free to do more of it and advance their field more than they would have otherwise. Whether the company in question is conducting scientific research, running a health care institution, writing software, or providing computing services for clients, all of these benefit from the mass availability of cheap, reliable open source infrastructure. Open source software is the great enabler of innovation in many fields, including, but not limited to, software. This is a direct result of the socialization of software costs.

While open source software is largely free of charge, there is a cost of paying engineers to write it. The big dirty secret about open source software is that many engineers who spend time writing and managing open source projects are paid to do so, whether they work for an independent software vendor, an IT department within a large company, a government agency, a scientific organization, or a non-profit. These institutions do this because of the economic benefits of participating – by distributing the costs around a large block of seemingly unrelated organizations, they all benefit by being able to use freely available open source software, relying on its low cost in order to run their operations more effeciently and economically. The myth of open source is that it's all written by kids in a basement or college students with nothing better to do with their time. Nothing could be further from the truth. There is a real economic benefit to participating in open source projects, and the smarter organizations view it as necessary overhead, because it sure beats the alternative of paying premium prices.

Example 1: Healthcare

When applying the lessons from open source infrastructure to the inner workings of government, one thing becomes clear – our infrastructure costs are way too high. One obvious example is health care. Our businesses are saddled with the rising cost of healthcare, which puts a damper on their ability to compete in a global marketplace. The reason businesses have to face this challenge is that, at least historically, Americans have resisted the idea of socialized medicine, refusing to pay the taxes required for such a system. The open source approach would be for the costs of medical care to be socialized and available to everyone at a nominal cost. The societal and economic benefits of such a system are readily apparent, for not only would more people be able to afford health care, but our businesses would be free to spend their dollars formerly reserved for employee health care on innovation. It would seem, then, that the traditional argument against socialism, that of preventing innovation, is turned on its head. Instead of socialized health care being an obstacle to innovation, it is a catalyst for more innovation.

Example 2: Telecommunications

Now let's consider the issue of our telecommunications providers and why fundamental reform is necessary to repair the economy. I will use as an example our subsidies for roads and highways. We don't expect to pay tolls for most of our roads. We expect our roads to be available, relatively free of problems, and a cheap way to get from point A to point B. The benefits of this are obvious – by subsidizing our highway system, the cost of transporting goods is significantly reduced, thereby freeing up capital that would otherwise be spent building roads and bridges. A similar tactic would yield similar results with telecommunications.

The state of our national telecom infrastructure is hardly becoming of an industrialized country, particularly the world's largest economy. By not pushing our telecom capabilities to at least match the levels of Korea, Japan, and several other industrialized nations, we are missing many opportunities to bolster our struggling economy and reestablish ourselves as a global technology leader. I call this, the “infrastructure gap.” Imagine a federal program to lead a nationwide effort to construct fiber optic cable plus all the necessary infrastructure pieces to reach every municipality, in the same way that electricity grids were constructed. What would that do to the price of bandwidth? Now consider how much our modern economy relies on the fast, efficient transfer of bits around the globe – let's face it, as a culture we're completely dependent on bits and bytes delivered via telecommunications infrastructure.

Socializing the cost of telecom infrastructure will have 2 clear benefits: there will be greater bandwidth available in places where it currently doesn't exist or is prohibitively expensive. Furthermore, with the drastically reduced cost of telecommunications services, businesses will have more capital available to invest. Given the recent developments around cloud computing, imagine the possibilities with cheaper, ubiquitous bandwidth. This would fuel a boom of more services delivered via the cloud in ways that aren't possible now, with 2 clear beneficiaries: entrepreneurs rushing to provide services via the cloud as well as entrepreneurs who build new business that take advantage of those cloud services. The latter would be able to make money off of services that cannot currently be efficiently automated.

With the Obama administration and the continued global economic crisis, there has been talk of a *new* new deal. Drawing from what we know about the proliferation of open source software, the new new deal should focus on programs to eliminate the infrastructure gap. Let's start with telecommunications - establish a nationwide effort to push our telecom infrastructure into the 21st century. Doing so will lead to a boom in new startups taking advantage of this technology and showing the way forward to future economic success.


Michael said...

John Mark,
Interesting analysis. The problem with calling more and more things "infrastructure" is that you start to get into the tricky business of getting funding to pay for building that infrastructure. As long as something is considered a product or a direct satisfier of "utility" people feel like they are supposed to pay for it, even though the may feel like it is a platform for products rather than a product itself. If you are saying that "no, people are not supposed to pay for it except through taxes (or charitable giving)", you are shutting off that activity from direct revenue generation.

We will need to see a sea change in the attitudes of Americans (and consumers around the world) in terms of their attitude about paying for infrastructure in order for the efficiencies of declaring stuff "infrastructure" to be realized. You are right to make a political analogy in the beginning..this is a highly political process. People will have to get straight about the following chain of ideas: "we like public goods", "we are willing to sacrifice money for public goods", and "look, here is my (big enough) payment for public goods".

Matt Asay said...

I'm not sure I'd call MySQL infrastructure, at least in the sense that there's somehow a public body either funding it or contributing to it. Neither is true. MySQL, the company (or, rather, Sun), writes virtually all of the database. If Sun went away, the "community" would languish, because the skills to write a good open-source database are not widespread.

So, Google and its friends love to treat MySQL as "infrastructure," but I'm sure Sun would prefer that they actually paid for the value they derive from it, such that Sun could write more great "infrastructure."