The more things change, the more they stay the same. Consider how changed a world we live in today when The Economist openly questions the bulk behavior of capitalists as evil bureaucratic rent-seekers and suggests that perhaps Karl Marx has something to teach after all. But the world remains stubbornly the same, as expert after supposed expert attempts to argue that open source software makes no economic sense and that a company like Red Hat cannot, therefore, exist (the latest example being this article on Medium.com).
W. Edwards Deming said "experience teaches nothing without theory," so I'm going to explain the theory that I believe underlies the 30+ years of experience I've witnessed in the world of successful open source software. A disclaimer: I didn't develop this theory. Credit goes to Ronald Coase (Nobel Prize in Economics, 1991), Oliver Williamson (Nobel Prize in Economics, 2009), and others. And indeed, I was unaware of this theory when I started Cygnus Support, the world's first company to provide commercial support for free software back in 1989. But I did joke, in all seriousness, that someday an economist would win the Nobel Prize in Economics for explaining the theoretical basis of that company. Open source exceeded expectations yet again when not one, but two economists were so honored. And so I begin with a lengthy paraphrase of Coase's Nobel Prize lecture to set up the theory.
A principal theme of Adam Smith's The Wealth of Nations was that government regulation or centralized planning are not necessary to make an economic system function in an orderly way. The economy could be coordinated by a system of prices (the "invisible hand") and, furthermore, with beneficial results. Lionel Robbins, one of Coase's teachers, raised the flag (as a "glaring deficiency") that (as Coase described), "an economist does not interest himself in the internal arrangements within organisations but only in what happens on the market, the purchase of factors of production and the sale of the goods that these factors produce. What happens in between the purchase of the factors of production and the sale of the goods that are produced by these factors is largely ignored."
Later in his Nobel Prize lecture, Coase said:
"The firm, in mainstream economic theory, has often been described as a 'black box.' ... This is very extraordinary given that most resources in a modern economic system are employed within firms, with how these resources are used dependent on administrative decisions and not directly on the operation of a market. Consequently, the efficiency of the economic system depends to a very considerable extent on how these organisations conduct their affairs, particularly, of course, the modern corporation. Even more surprising, given their interest in the pricing system, is the neglect of the market or more specifically the institutional arrangements which govern the process of exchange. As these institutional arrangements determine to a large extent what is produced, what we have is a very incomplete theory."
Coase decided to study the vertical and lateral integration of industries in the United States. British economist Arnold Plant described in his lectures the different ways in which various industries were organized, but it seemed to lack any theory that would explain these differences. Coase set out to find it, relaying in his Nobel lecture:
"Plant was opposed to all schemes, then very fashionable during the Great Depression, for the co-ordination of industrial production by some form of planning. Competition, according to Plant, acting through a system of prices, would do all the co-ordination necessary. And yet we had a factor of production, management, whose function was to co-ordinate. Why was it needed if the pricing system provided all the co-ordination necessary?"
"I found the answer by the summer of 1932. It was to realise that there were costs of using the pricing mechanism. What the prices are has to be discovered. There are negotiations to be undertaken, contracts have to be drawn up, inspections have to be made, arrangements have to be made to settle disputes, and so on. These costs have come to be known as transaction costs. Their existence implies that methods of co-ordination, alternative to the market, which are themselves costly and in various ways imperfect, may nonetheless be preferable to relying on the pricing mechanism, the only method of co-ordination normally analysed by economists. It was avoidance of the costs of carrying out transactions through the market that could explain the existence of the firm in which the allocation of factors came about as a result of administrative decisions (and I thought it did). In my 1937 article, I argued that in a competitive system there would be an optimum of planning since a firm, that little planned society, could only continue to exist if it performed its co-ordination function at a lower cost than would be incurred if it were achieved by means of market transactions and also at a lower cost than this same function could be performed by another firm. To have an efficient economic system it is necessary not only to have markets but also areas of planning within organizations of the appropriate size. What this mix should be we find as a result of competition."
When Cygnus announced its corporate tagline, "We make free software affordable," we were realizing the thesis of Ronald Coase, although we did not know it. We were establishing a market price for all the planning that needed to be done to identify, collect, configure, test, document, release, and support packages of free software competitive with proprietary software. And with the benefit of hindsight, the record shows that we intuited our way into Coase's theory just as his work was receiving the Nobel Prize. As I recount from my chapter of the Open Sources book:
"In 1991, we hired a young business student, recently laid off from Applied Materials, who wanted to learn how to sell software. Though her native language was not English, she picked things up very quickly. By no means a hacker (though she spent some weekends at Cygnus teaching herself to program in C), she nevertheless became a really strong advocate of the open source approach. After six months of very successful selling, she invited me to see her make a customer presentation. I was floored. I had always sold open source the way a hacker would sell it: focusing mainly on the technical merits. She explained the intrinsic complexity of the job we were doing and the business value of the software we delivered, and this helped us finally explain to customers why they should buy from us instead of trying to do the work with their own people. I was selling the fact that our engineers were somehow better than theirs (not a message most managers want to hear), whereas she could explain how their engineers would benefit from having us do the baseline porting, support, and maintenance work."
Before Coase's theory was accepted, the premise of a competitive firm rested on control of input costs and/or access (usually through exclusive property title or mineral rights), control of labor costs (historically often through coercive means), and occasionally patents and copyrights that prevent others from practicing or reproducing whatever it is that is claimed by the patent or copyright, and ultimately the competitive position that those three control points afforded the resulting product or service. But after Coase, it was theoretically possible that if input costs went to zero, if labor costs effectively went to zero, and if there were no patents or copyrights preventing unlimited duplication of the outputs, there would still be transaction costs that could be optimized and a superior competitive position to be obtained. This is the jump to light-speed that Cygnus made in its early days, and that Red Hat has managed to do at enterprise scale.
Successful open source software companies "discover" markets where transaction costs far outweigh all other costs, outcompete the proprietary alternatives for all the good reasons that even the economic nay-sayers already concede (e.g., open source is simply a better development model to create and maintain higher-quality, more rapidly innovative software than the finite limits of proprietary software), and then—and this is the important bit—help clients achieve strategic objectives using open source as a platform for their own innovation. With open source, better/faster/cheaper by itself is available for the low, low price of zero dollars.
As an open source company, we don't cry about that. Instead, we look at how open source might create a new inflection point that fundamentally changes the economics of existing markets or how it might create entirely new and more valuable markets.
Let's look at a specific example. On October 16, 2009, the CIO of the US Department of Defense issued guidance on the use of open source software that began:
"To effectively achieve its missions, the Department of Defense must develop and update its software-based capabilities faster than ever, to anticipate new threats and respond to continuously changing requirements. The use of Open Source Software (OSS) can provide advantages in this regard."
This is really quite a revolutionary statement when one considers how rule-bound the federal acquisition process is and how much it requires concrete specificity before any bid can be written or reviewed. I don't want to distract the reader too much with stories about how such rules led to a 57-cent screw costing $76 or how a $6 hammer winds up costing over $400. Rather I want to emphasize that, notwithstanding the government's best efforts to negotiate the best deal for everything they buy, those rules were thought to prohibit them from buying anything they could not even describe. Responding to continuously changing requirements is a fundamental paradigm shift.
Continuously changing requirements take us beyond simple markets and prices and into a world that includes governance. Oliver Williamson set it up for us in his Nobel Prize lecture:
"The research program on which I and others have been working has been variously described as the 'economics of governance,' the 'economics of organization,' and 'transaction cost economics.' As discussed in Section 1, governance is the overarching concept and transaction cost economics is the means by which to breathe operational content into governance and organization. The specific issue that drew me into this research project was the puzzle posed by Ronald Coase in 1937: What efficiency factors determine when a firm produces a good or service to its own needs rather than outsource?"
Or, to the point of our initial question: What efficiency factors determine when a firm should produce its own open source capabilities to meet its own needs rather than outsource? And to the point of the real question: What is the optimum of planning that maximizes competitive benefit when one has a perfectly continuous choice of buy-vs-build when it comes to software?
Imagine one runs an international airline company. Or a stock exchange. Or a manufacturing facility. Or a project for the Department of Defense. Technology is certainly a requirement for basic operations. But unlike the cost of a gallon of fuel, a square foot of real estate, or a person-hour of labor, technology raises major governance issues the instant one looks beyond what should work today to what may need to work efficiently and effectively tomorrow. Alas, the history of enterprise information technology is not a very pretty one.
As Red Hat CEO Jim Whitehurst reminds us in The Open Organization, companies spend a lot of money on software that never delivers its intended value. About 18%–23% of all IT spending goes into projects that are abandoned before they go into production. Another 55%–65% is spent on software that is late—sometime very late, lacking key functionality, or with operational defects that materially affect performance.
I have looked at these numbers for a number of years, and am quite confident in the estimate that at least 30 cents of every dollar spent on IT is wasted due to the poor performance that results from the perverse incentives of proprietary software. In aggregate, enterprise IT buyers are wasting hundreds of billions of dollars every year. Worse, the ROI they are failing to deliver (which should be six to eight times their IT investment) adds up to trillions of dollars every year. The buying patterns of most companies are still geared toward buying concrete capabilities that militate against transformative innovation rather than encouraging and supporting it. Which is a colossal failure of governance and economics.
As I have already stated, open source as a commodity replacement for proprietary software is an economic path to zero price and zero cost. That is still globally more efficient than spending positive dollars on software that often delivers negative value. But that's not a business model that leads to success or scale for the open source developer.
But open source as a platform for innovation and transformation offers huge value to the customer and a compelling value proposition for the vendor. To make money this way, the open source vendor must provide not merely a feature that signifies or demonstrates an innovative capability; it must align with and support the governance goals of its clients. It must make it possible for the client to innovate in ways not possible with legacy proprietary software or the churn and uncertainty of in-house development.
That's non-trivial, and perhaps more fitting to answer in a separate article than to elaborate further in this one. But it is possible. And it is also necessary, because proprietary software is unsustainable, both for those who write it and especially for those who buy it. It is all the more necessary because as long as our economy relies on growth while being limited by finite resources, innovation is the only sustainable approach to long-term growth. Managed effectively (i.e., with an optimum of planning), open source can lower barriers to innovation and increase the effectiveness of collaboration, enabling continued positive economic growth. How to do this is explained in detail in The Open Organization.
Next year, Coase's first transmissions will be four score and seven years old. That should be enough time for the average person with above-average intelligence to read them, understand them, and to approach the future ready to apply them productively. To paraphrase an old open source maxim, if you can't solve the problem, please get out of the way of those who are presently solving it.