Recently I was invited to comment on the following question: Is it possible to decouple values from the global standardization of critical and emerging technologies (CETs) such as 5G and AI, given that such technologies do involve national security and ethical issues?
The above question is dependent upon a series of implicit assumptions which make up the current standardization environment.These assumptions are rarely made explicit because standards and standardization are so ingrained that people forget to question them. As an example, everyone “knows” that traffic signals use red/yellow/green (usually top to bottom) – it is nearly universal. But why is it this way? Because the earliest accepted practice for wiring was red for a hot wire, green for a ground wire. This color scheme was picked up by railroads for their signalling, and then moved to motorways in the US (apparently starting with the Detroit police department). It is now a standard – common across nearly all countries. Is it still the best system? Possibly not, just like QWERTY is not the best keyboard layout. But this then gets into a discussion of path dependencies and implications of standardization.
It is why, as with any simple question, the complexities only become apparent when the issue is discussed at a detailed level. It is the intent of this micro-treatise to look at some of the hidden complexities of standardization and to try to reframe them for a larger discussion (as a caveat, during my career in standardization, I tried to be an embedded, neutral, and empowered participant when I had the luxury of not having to defend a corporate position).
The hidden complexities of standardization
To clarify, a standard is a technical document that specifies a technical parameter that allows duplication or interconnection. It is provided so that others can share in the use of the product, service, or other described technology. It is morally neutral, is neither wrong or right, and, in and of itself, is a benign document. The same is not true of standardization, the discipline that creates standards.
The issue comes with “standardization”, the act of creating and using the document, and this is where the conflicts arise. As a bit of background, the current standardization regime – typified by ISO, IEC, ITU, and WTO – was designed after WWII for an industrial society, when industry was capital intensive, and participation was difficult except for a few large companies and countries. The post WWII framework was created to ensure that all countries that mattered (read countries that could cause a major war) had a stake in ensuring the common good.
For large capital goods providers from major developed countries, this scheme worked. For countries without a significant technical base, or countries who were agricultural, or countries who were moving up the industrialized chain, this scheme (the three-letter standardization philosophy) was less favorable. With no technical base, no IPR, and no economic influence, they were tied into a program that favored wealthy and established players.
This scheme, as I noted, was designed for capital intensive industries with limited players. In the post-industrial society, or maybe the massively networked society, the current standardization ecosystem can appear to be regressive and repressive. Rather than a small cadre of companies designing and producing products and services that are stable (think electric grid, roads, dams, machinery), you have thousands and thousands of organizations producing transient, short lived products and services, where “first to market” is a goal and necessity. IPR rules and roles are confused, and precedent and formal structure are treated with disdain because they serve no purpose except to empower existing social, economic, and political structure.
To use a dated example, the DVD players that were popular prior to the advent of streaming contained hundreds of patents, all of which had to be licensed to be used. The world’s largest manufacturer of DVD players were Chinese smaller manufacturers, who had to use the DVD standards to make products that could read DVDs. To implement the standards, a royalty payment of 4% of purchase price or $6 was required. This wasn’t too bad when prices for a player were hundreds of dollars; when the market drove the price to $40, these smaller manufacturers were effectively either closed out or bankrupted.
This is one of those paradoxes which are unresolved in the current standardization regime. Ideally, standards are supposed to “lift all boats” – that is, they should allow technology transfer between those who have it and those who need it. But the current international ecosystem allows a standard to contain patents which must be licensed. This isn’t bad if you are a large or wealthy corporate entity that has its own selections of patents to trade – a quid pro quo. But in the case of the DVD player manufacturers, there was no quid pro quo – only royalty payments. They couldn’t be avoided – laws in most of the importing countries required that IP rights be respected. In this case, the standards regime failed in its mission to transfer technology fairly; rather, it forced a wealth transfer in exchange for risky market access.
When the Internet took off and the World Wide Web came into existence, the entire standardization world started to change. For the software that was standardized by the W3C (the organization that is dominant in Web standards matters), royalty free standards became a mantra. At the same time, Open-Source software appeared, freeing much of the web’s software from the problem of royalties. But the hardware that drives the software is still encumbered, creating interesting approaches to computing.
Complicating matters, the divide has been largely ignored in other disciplines. Pharmacy standards and IP regimes favor the larger countries – the disaster that is COVID brought this issue to the fore. Similarly, IP rules for agriculture – while not embodied in standards but rather in standardized practice – are also becoming more and more divisive. Again, it is not the “standard” per se that is malicious – it is the implementation that creates inequities.
To complicate matters further, there is a rising tension in standardization because the utility and needs for standardization have become so necessary. I wrote once that “standardization is the hallmark of an industrialized society”. As society becomes increasingly interconnected and increasingly interdependent, the importance and ubiquity of standardization increase. But because the process of standards creation never been defined for an international networked society, and because the mechanisms for their creation have lagged current practice, standards, and standardization, as currently practiced, have become anachronistic. The internet – despite efforts to the contrary – does not stop at a border. Telephony no longer respects geographic divisions, and information and data seem to flow everywhere constantly, without regard to artificial boundaries. Yet most standards are created by national standards bodies operating in an international organization.[i] This is another source of conflict within the standardization ecosystem.
The nature of standardization
There are several other issues with standardization that are critical to its practice. The first of these is path dependence, which is rarely recognized except in the case of a real blunder. The most well-known misstep was the electrical outlet, of which there are fifteen variants. The issue of path dependence enters here because each country now cannot afford to change its plug without incurring significant costs, both in appliance and socket replacements. The most studied standards path dependency of which I’m aware is the four feet, eight ½ inch railway gauge[ii], which, despite years of existence, continues to be violated by various nations, all of whom have excellent reasons to incur the economic disadvantage of standardization. A second issue is control of the process. Currently, nearly all standards creation (both organizations and participants) is funded by manufacturers, who have no incentive or desire to change the system.
Standardization is, above all, a tool to create, retain, and control markets. It is currently managed by corporate or business practitioners in the industrialized world. In the US and Europe, it has been argued that industry has come to control standardization and standards (and hence the future of progress because of the issues of path dependency.) Schoechle (Standardization and the Digital Enclosure), Busch (Standards: recipes for reality),MacKinnon (Consent of the Networked), and even Veblen (Imperial Germany and the Industrial Revolution) all point out the power of standards and industry ownership (although Veblen may be a stretch). Nations who came later to the party (especially China) are handicapped by what I call the G7 standardization regime, which is carefully crafted to empower the G7. This would normally lead to a Balkanization of standards, except that this would raise product costs or limit markets for manufacturers.
The central issue that has not been discussed is “What is standardization”? In the US and partially the EU, standardization is a private sector activity given limited immunity from anti-trust by use of rules and structure. China’s newer standards regime makes standardization more of an industrial policy tool that has a more governmentally managed aspect. The nature of standardization has never been made explicit, however. I would argue that standards are impure public goods, given their nature of being limited in number (practically) but serving a public good (interconnection and interdependence). The implication of this, of course, is that an impure public good can be legislated, because control of an impure public good has implications for the body politic.
To expand on this idea, it is reasonably certain that a standard confers economic advantage, especially if you hold the IPR in a widely needed standard. Additionally, the choice of a standard has path dependency issues that should be considered. As an example, if the standard requires use of an ingredient which requires destruction of an environmental element (land, mineral resource, increased pollution), then there is a responsibility for governmental oversight. In many countries, there is a requirement for an environmental impact statement prior to approval for building. We’ve seen how standards can have an impact on a society, culture, or economy, yet there is no “future focused study or examination” required to create, implement, or manage one. Yet, because the extant standardization regime is driven by and for corporate interests, there is a sense of laissez-faire that pervades thinking on this. The attitude is “don’t break any unfair competition or anti-monopoly rules and you’ll be fine.”
There is an unwillingness to recognize that standardization and standards are powerful economic tools that can be used by governments. A standard is a codification of a business practice expressed in technical terms. The business practice has social, political, and economic bias, and because the standard reflects these biases, it inculcates the values of the creators. As an example, privacy on the web can be achieved in a much more widespread way with the inclusion of additional technologies. However, so doing would cripple tracking and web marketing, making it difficult for many marketing and advertising companies to target ads. If a standard is created at the behest of the advertisers that makes it difficult (or more difficult) to invoke privacy protection, it would be completely acceptable (by definition) if it met the process requirements of the creating organization. It would, however, probably be socially unacceptable – but that is not taken into the creation equation.
So, standards reflect the bias and desires of their creators, within the constraints of the organizations that create them. Currently, as far as I know, there are no environmental impact, future economic constraints, or potential social disruption studies required, yet these are all known results of standardization, and which are governmentally regulated in other cases. You cannot decouple the social, political, and ethical issues of a standard (and the technology it standardizes) from the standardization organization that created it any more than you can disconnect the laws and regulations of a country from the legislature that created them.
Is standardization an industrial policy tool, a social change tool, a legal issue, an economic issue or what? It is all things to all people, governments, organizations, and companies. But industry has clothed it technospeak so that it will not be regulated.
My proposal would be for at least one government to create a “Department of Standardization” at a cabinet or equivalent level. It is a cross discipline function that would impact other governmental functions (education, environment, commerce, defense, natural resources, and so on). Its responsibility would be to oversee the creation of an open process that does not limit outcomes, but rather ensure that input considers all stakeholders and stakeholder positions in an equitable fashion. The success of the function (judged by national economic success) might inspire others to copy or make similar adjustments to their national infrastructure. It might, although this is Pandora speaking, even result in a realistic study of standardization as a discipline in its own right.
As a bit of background, my initial field of study in college was Medieval European history. I was especially intrigued by the Investiture Controversy, a centuries long dispute on ecclesiastical versus secular authority. It was finally solved when the disputing sides were forced to document and establish their positions, which lead, indirectly, to the modern European nation state and “sort of” a solution to the issue of the legitimization of authority.
About the author
Carl Cargill has been involved in standardization for forty years, primarily participating in or managing standardization activities for several technology companies, including Digital Equipment Corporation, Sun Microsystems, Netscape, and Adobe. He has written several books, numerous articles, testified before the US Congress, and served on standardization committees and standards organization management committees.
[i] Since 1980, consortia (a variant method of producing specifications) have become common in standardization, producing specifications for use by industry. Legitimized in the US by an Act of Congress, they provide an alternative to international standardization organizations. There are now over one thousand consortia busy writing “standardized specifications”. See https://www.consortiuminfo.org/sso-list/, a comprehensive list published by Gesmer Updegrove LLC, probably the best source for information on this form of standardization. The Chinese government has recognized “social organizations” which serve similar function for China.
[ii] See Tracks across Continents, Paths through History: The Economic Dynamics of Standardization in Railway Gauge by Douglas J. Puffert, an interesting and enjoyable study of the issue.