In March 2009, I wrote a series of blogs Concentration at the ODF TC, Concentrate! 10:1 and Another leap forward for openness in which I called for broader representation and more participation by non-vendors at the OASIS-Open ODF TC (with a similar principle of balance required for the OOXML and other standards groups.)
These followed up on a point I made in 2007's Is our idea of "Open Standards" good enough? verifiable vendor neutrality and followed up in 2008 Neelie Kroes gets it: "Standards are the fountain of intereoperability" that "openness" as defined as being potentially open for anyone to participate may not be adequate as a basis for mandated standards, due to the cartelization issue.
So a particular phrase in the US Office of Management and Budget Circulare A-119 Federal Participation in the Development and Use of Voluntary Consensus Standards and in Conformity Assessment Activities (from 1998, hat-tip Pattrick Gannon on the OpenDocument users list) stood out (I put it in bold):
A voluntary consensus standards body is defined by the following attributes:
(ii) Balance of interest.
(iii) Due process.
(vi) An appeals process.
This balance of interest can only come out of broad representation: indeed, they are two sides of the same coin.
Openness gives the potential for a balance of interest, but it does not guarantee it. The participation of all significant vendors in a committee similarly gives the potential for a balance of interest, but cannot deliver it without adequate participation by users.
An open standard developed with only vendor involvement may help define a market, but it does so in a heavy-handed way that actually circumvents the market. The market is not only a mechanism for establishing supply and demand and price points; it is also, when there is enough substitutabilty and knowledge and rationality, a mechanism whereby demand feeds back to suppliers information on the mix of features required.
This kind of feedback is an important part of the operation of the market when there are technologies that have not stabilized. In this situation, an open standard developed with vendor-side dominance actually retards and impedes the market.
Lets take a butcher's market for example. I don't suppose that many of us live in place where such a thing exists now, for example a town's shopping street with several different butcher shops. There is a certain freedom from choice and easier comparison if all the butchers decide to provide only the same cut of meat: you want lamb chops, you can have lamb chops that are 1 cm thick, that is the standard.
But part of the operation of the market is to allow consumer preference to express themselves: the big fat butcher has 1cm chops but the skinny small butcher next door has 1.5cm chops. This year thicker chops are the fashion, so people buy from the skinny small butcher, and the big fat butcher must decide whether to make his chops thicker to catch that market or the same to serve his loyal customers or even thinner to differentiate more.
A standards body with a balance of interest from broad representation acts more like this competitive market: the standard takes care of issues further along in the market negotiation. Rather than an agreement by suppliers on what to offer, it is more an agreement on what will be offered and accepted.
UPDATE: What does ANSI Essential Requirements actually say?
Readers amazed at the exchange below may, in the interests of woeful inadequacy, be keen to see for themselves what ANSI's rules for balance are for developers of its standards. They are in the document Essential Requirements.
There we find very clear statements on balance:
The standards development process should have a balance of interests. Participants from diverse interest categories shall be sought with the objective of achieving balance.
and in the Benchmarks section
Historically the criteria for balance are that a) no single interest category constitutes more than one-third of the membership of a consensus body dealing with safety-related standards or b) no single interest category constitutes a majority of the membership of a consensus body dealing with other than safety-related standards.
The interest categories appropriate to the development of consensus in any given standards activity are a function of the nature of the standards being developed. Interest categories shall be discretely defined, cover all materially affected parties and differentiate each category from the other categories. Such definitions shall be available upon request. In defining the interest categories appropriate to a standards activity, consideration shall be given to at least the following:
c) general interest.
Where appropriate, additional interest categories should be considered.2
Appropriate, representative user views shall be actively sought and fully considered in standards activities. Whenever possible, user participants shall be those with the requisite technical knowledge, but other users may also participate. User participation should come from both individuals and representatives of organized groups. There are several user categories:
1. User-consumer: Where the standards activity in question deals with a consumer product, such as lawn mowers or aerosol sprays, an appropriate consumer participant's view is considered to be synonymous with that of the individual user - a person using goods and services rather than producing or selling them.
2. User-industrial: Where the standards activity in question deals with an industrial product, such as steel or insulation used in transformers, an appropriate user participant is the industrial user of the product.
3. User-government: Where the standards activity in question is likely to result in a standard that may become the basis for government agency procurement, an appropriate user participant is the representative of that government agency.
4. User-labor: Where the standards activity in question deals with subjects of special interest to the American worker, such as products used in the workplace, an appropriate user participant is a representative of labor.
1 "Affiliation" refers to the entity that the consensus body member represents (which may or may not be that person's employer). If the consensus body member is serving in an individual capacity, then the name of the individual, that person's employer, sponsor and interest category should be available. Contact information is not required.
2 Further interest categories that may be used to categorize directly and materially affected persons consist of, but are not limited to, the following: a) Consumer; b) Directly affected public; c) Distributor and retailer; d) Industrial/commercial; e) Insurance; f) Labor; g) Manufacturer; h) Professional society; I) Regulatory agency; j) Testing laboratory; k) Trade association.
Hmmm, a majority of the membership...sound like what I have been suggesting, doesn't it. If it is a good idea for ANSI standards—indeed an essential requirement—, why is it not also not a good idea for OASIS in general for its TCs? And for TC members to pro-actively show their good citizenship and adhere to? These requirements are essential to ANSI for good reasons: because of the legal obligation to block cartelization and undue dominance, because they are inimical to the standards system. Quibbling doesn't change that.
And the same thing goes for the ECMA TC45 and SC34 WG4, by the way, in my book. The difference is that I think that MS pretty much groks that it is precisely having a balance of interests at TC45 (and SC34 and WG4 and WG5) that will allow IS29500 to have credibility not just as an "open standard" but this higher bar as set by ANSI and other NBs. This is particularly true if IS29500 moves from being the backwards-looking standard to being the forwards-looking one.
On a tangent, it is interesting that the very next section of that ANSI Essential Requirements brings up the old chestnut of standards which conflict. Readers with sad lives may remember how there were claims made during the first round of the OOXML standardization process that it contradicted ODF because it was in the same area: at the time I was puzzled that certain people would just make things up about how standards work, without bothering to look or to ask old-timers. So what is the current position (well, 2008)?
2.4.1 Definition of Conflict
Conflict within the ANS process refers to a situation where, viewed from the perspective of a future implementer, the terms of one standard are inconsistent or incompatible with the terms of the other standard such that implementation of one standard under terms allowable under that standard would preclude proper implementation of the other standard in accordance with its terms.