Creating A New Kind of Public/Private Partnership

for O*NET

 Prepared for:
U.S. Department of Labor
Office of Policy and Research
200 Constitution Avenue, NW
Washington , DC 20210

 The views expressed in this paper are those of the author only and do not necessarily represent the views of the Department of Labor or its contractors.

Executive Summary

Government is finding new ways to get its work done. While some changes can be incremental, others require something more radical.

The following paper describes a challenge – the computerization of the Department of Labor's Dictionary of Occupational Titles – that requires government to stretch beyond its current notions of public/private partnerships and break new ground.

Mr. Getzendanner suggests that business and government become co-owners of a new kind of organization dedicated to the development and growth of a whole industry – the development and delivery of labor market software. Its purpose would be to create a system capable of delivering a premium selection of tools and information resources into every labor market in the country.

Unlike traditional trade associations or collaborations, this organization would actually be the infrastructure around which the industry organizes itself and pursues both its public and private goals. Government would be free to target its resources wherever the greatest public good could be achieved, or it could choose to withdraw completely. Both the public and private sectors would see an explosive increase in their ability to achieve their respective labor market goals. While the organization would exhibit an interesting blend of the principles of Jeffersonian democracy, Reinventing Government, and the Internet, the design presupposes no specific political or economic orientation.

Based heavily on the pioneering work done at Visa, the credit card company which chose a similar organizational strategy, Mr. Getzendanner outlines the design process that a group of early leaders would need to go through to create such an entity. He then posits a business function that could serve as the heart of such an effort – a "labor market software registry" – and outlines some of its likely features. Finally, he describes possible next steps and points to other similar challenges – such as governing the Information Superhighway – that are likely to benefit from the experience of this effort.

Section 1: O*NET and Visa

All truly creative leaps are mind-boggling before they are achieved, and mundane immediately afterwards. They become the building blocks for everything that follows – so ingrained in the day-to-day grind of life that nothing else seems imaginable. Yet beforehand, they are laughable oddities. So it has been with gadgets like the telephone or the computer. So it has been with grand ideas, such as democracy.

So it will be with the next generation of public/private partnerships. The world will change when something profoundly new and profoundly right comes along.

Today, we argue whether a program or function should stay totally within the current public realm (government), move totally within the current private realm (privatization) or become some carefully crafted blend of the two (partnership). These might be the right questions for existing programs or services, but it is less obvious when considering something that is itself a new kind of challenge – such as managing the Internet – where the old rules don't seem to apply.

New challenges require new responses. Those who successfully lead this charge into the future will be doing a great public service, and are likely to gain enormous political and economic advantages, as well.

A Prime Opportunity – O*NET

For several years, DOL has struggled with an aging Dictionary of Occupational Titles. The Dictionary has provided the basic language for people who want to talk about jobs in the economy, generate employment statistics and refine training programs to prepare workers for those jobs. But times have changed. The economy has become enormously complex and the nature of its jobs is changing constantly.

O*NET is DOL's effort to update and computerize the Dictionary, using all of the technological know-how it can organize. Although it is only one initiative among several at DOL that is trying to fashion a comprehensive labor market information system, it is a pivotal one. It's purpose is to construct a shared foundation of language and analysis, just like the Dictionary before it, on which both public and private efforts could build. Without something like O*NET, the labor market system does not become a meaningful whole, but remains just a series of disconnected fragments.

Still, no simple classification scheme of jobs and skills like the Dictionary will suffice any longer. No single institution – be it public or private – can organize all the information that is needed and keep it updated. The economy is in the midst of a fundamental transformation. "High performance work organization," virtual corporations," and "learning organization" are just a few of the numerous phrases meant to describe the leading edge of that change. But the transformation is not yet completed. It is only beginning. There is no way to know where or when it will end. The "labor market" that a public system needs to support is not only new, it is still in formation.

Although electronic technology makes possible a number of new ways to organize information and can create new pathways for labor market participants to communicate with each other, creating a system that can truly be as complex and dynamic as the labor market is a daunting task. The primary stumbling block is not technological, however, it is institutional.

The situation requires a new kind of organizational structure: something as new as the economy it hopes to serve. For example, the old Dictionary allowed an arms-length interaction with business. A new system would not only need businesses' intimate involvement, it would need to become a integral part of businesses' day-to-day operations. This is the only way for such a system to keep up with the rapid changes in the economy. Then, given the need to protect sensitive proprietary information, such a system would need to have a distinctly new ownership structure, a distinctly new governance structure, and a distinctly new way to grow and reach the necessary scale. The structure must be absolutely trustworthy to all those involved. No existing public/private partnership serves as an adequate prototype.

A Powerful Prototype – Visa

When venturing into new territory, it is always helpful to have a roadmap prepared by someone who has gone there before.

In the late 1960's, a group of banks faced the challenge of constructing a system to exchange credit card transactions. The coming age of electronic transactions was still on the horizon, and they needed a system that could let them exploit it when it came. Since no single bank could afford creating the whole system, they had to create an organization that they could all collectively own, all individually trust, yet would not impede competition. In fact, it would need to enhance competition if it hoped to pass antitrust muster. No organizational structure existed that met these criteria. They had to invent it from scratch.

Led by the credit card manager of a small regional bank, Dee Hock, a small group of bankers began working on a core set of purposes and principles to which they believed all banks would agree. They then steadily involved more and more banks in a process of refining that core set of beliefs about their industry, and began to conceptualize a possible function and initial structure for an organization.

After a year and a half of intense effort, the process was complete and the organization that would become known as Visa came into existence with Mr. Hock as its first CEO. One of the great ironies was that many of the principles they agreed to would be recognized by historians as parallel to those at the heart of the U.S. Constitution. The design process had given rise to a corporation unlike any other. In 1993, Mr. Hock, who retired from Visa in the mid-1980s, described Visa in the following way (used with permission):

I still find it difficult to describe the organization but let me try.

In the legal sense, VISA is a non-stock, for-profit, membership corporation. In another sense, it is an inside out holding company in that it does not hold but is held by its functioning parts.

The financial institutions which create its products are, at one and the same time, its owners, its members, its customers, its subjects and its superiors.

It is an organization, the totality of which, excluding thousands of affiliated entities, would, if converted to a stock company, have a market value in excess of $70 billion. Yet it cannot be bought, traded, raided or sold since membership is held in the form of perpetual, non-transferable membership rights. However, that portion of the business created by each member is owned solely by them, is reflected in their stock prices and can be sold to any other member or any entity eligible for membership, a very broad, active market.

A staff of less than two thousand scattered in twenty one offices in 13 countries on four continents guides this half trillion dollar business. It has multiple boards of directors within a single, legal entity, none of which can be considered superior or inferior, as each has irrevocable authority and autonomy over different geographic or functional areas. Their interests are often opposed yet they are served by common officers and staff.

It espouses no political, economic, social or legal theory, transcending language, currency and culture to successfully work with and harmonize relations between countries, institutions and peoples of every persuasion. It has gone through a number of wars and revolutions, the belligerents continuing to share common ownership and never ceasing reciprocal acceptance of cards or exchange of sales drafts even though they were killing one another.

It is an organization which, in five years, transformed a troubled product with a minority market share into a dominant market share and, by a substantial margin, the most profitable banking service. It has had no less than twenty and as much as fifty percent compound annual growth, through the best and the worst of times.

Its product is the most universally used and recognized in the world, yet the organization is so transparent its ultimate customers, most of its affiliates and some of its members do not know it exists or how it functions. At the same time, the central core of the enterprise has no knowledge of, information about or authority over a vast number of the constituent parts. No part knows the whole, the whole does not know all the parts and none has any need to. The entirety is self regulating.

The process that Mr. Hock undertook (see Appendix I) is deceptively simple in concept and extremely difficult to sustain in practice. First, the banks needed to agree on a purpose for the organization. To do so, they had to "peel back the onion" to determine what the essential function of banking was, how it related to other core concepts, such as "money," and what was changing in the marketplace that was making the old ways of doing business obsolete.

Second, they clarified the principles by which an organization could be brought into existence to meet that purpose. These principles were more than just platitudes, they were specific criteria of design that could not be violated in pursuit of the purpose. There could be no compromise on any of the principles.

Finally, they came up with a concrete concept for an organization and proposed an initial structure and set of membership rights from which whole organization would eventually grow.

Applying the Visa Process to O*NET

Obviously, Visa is not a direct model for O*NET. Still, Visa needed to construct a new computerized system where none existed before. It needed to draw in the voluntary participation of commercial enterprises. It was determined to construct its structure on what amounted to Jeffersonian principles of openness and self-governance. The parallels are tantalizing.

But like Visa, conceptualizing an organization for O*NET requires starting from scratch.

The remainder of this paper takes Visa's approach of starting with the broadest and most complete set of goals upon which there can be agreement, and then steadily refining them into progressively more tangible design rules for an organization. Each subsection takes this process one step further until an initial concept of an organization can be found.

Section 2: From Purpose to Concept

A First Cut at Purpose – Broad Goals and Functions

Policy makers have spent a lot of time talking about the purposes of public systems. Labor market systems are no exception. And while there is considerable disagreement about the level of governmental involvement in the functioning of labor markets, there is broad agreement that the ideal labor market system would do the following things exceptionally well:

• A system through which businesses could easily express their demand for labor, analyze their current and future needs, contract for educational services, construct their own tailor-made human resource applications, and offer their own labor market know-how for sale.

• A system through which individuals– using whatever source they feel is most trustworthy – could learn about economic opportunities, could learn what their companies wanted of them, and could craft their own development plans.

• A system through which educational institutions could develop and deliver curriculum, and interact easily with businesses and potential students.

• A system through which human resource service providers – public or private – could build unique and powerful products and practices.

The term "system" is being used to imply that there is a recognizable coherence to the whole, but it shouldn't be taken to imply any specific form or that it must be "directed" from any single place. A technocrat may dream of a top-down system in which every piece is controlled by a piece hierarchically above it and all driven by a single, uniform standard.

Certainly there are many people inside and outside of government that share this dream and are pursuing it aggressively. But this is unlikely to be practical. At the other extreme, a laissez faire libertarian may dream of no structure other than that which spontaneously emerges from the marketplace.

The current system is a mixture of these two extremes, and it is likely that an "ideal" system would be a combination as well. There is no question, for example, that market forces must be exploited to do most of the "heavy lifting," especially in this age of government cutbacks. However, standards are clearly essential to the functioning of markets, and while they sometimes spring up spontaneously, often a conscious and collaborative effort on the part of the market participants is necessary. "Public" standards often lead to the rapid acceleration of the development of markets. Getting one extreme to the exclusion of the other results in enormous waste. Both are needed, not "either/or."

The bottom line is that all participants in labor markets – businesses, individuals, and service providers – should have a vehicle through which to pursue their goals and interact with each other as productively as possible. This is a social function that has existed in numerous forms throughout recorded history and will exist in many more forms far into the future. We are just trying to find the form that will work especially well for the next generation or two.

A Key Underlying Change: One Size Does Not Fit All

In a world where the notion of constant change is long since a cliché, nothing is changing faster than the language of labor markets. Every new piece of technology adds new words and standards. Every new "autonomous team" develops a unique language and ethic about the content and process of its work. Every new strategic alliance among firms requires them to meld the languages they have developed independently. And these changes are not reaching a plateau, they're accelerating with no end in sight.

Faced with this " Tower of Babel" within our labor markets, one's first reaction might be to try to create a single, common language that would please the technocrat. Yet this multitude of languages is emerging precisely because different demands are

being placed on workers in different economic circumstances. The more advanced one's workforce strategy, the more advanced one's language must be to carry it out. And in these changing times, a language that is perfect today will have to be replaced tomorrow. One size does not fit all, and it doesn't fit for long.

Still, the cost of being the only one to use a given labor market language is enormously high. If a firm is the first to enter new economic territory, the expense may be worth it. But normally it is better to share a language with as many others as possible. Costs are lower. The market of possible suppliers, partners and customers expands proportionately. While a firm may want to create a unique "dialect" that corresponds to its specific niche within a market, there are enormous incentives for a firm to join the crowd – or to convince the crowd to follow its lead.

So, both standards and differentiation are required, both rapid adaptation and stable platforms from which to operate are needed, both aggressive competition and collaboration are a necessity. It must be both public and private. Again, either/or will not suffice.

While this duality has always existed in the economy, we are coming out of a period where mass production of standardized products was the norm. The Dictionary of Occupational Titles worked well because jobs tended to be standardized, too. The system matched the economy of its time. A new system must be created to match the new economy.

A First Cut at Organizational Principles

Of course, these pressures are not unique to labor markets and their supporting enterprises. Institutions of all types, both public and private, are struggling to make their way. Many have had to totally remake themselves, some several times over, so the popular business press is full of descriptions of what the successful organization of the 21st Century will be like. It is a minor matter to translate these characteristics into a description of an ideal labor market system:

• Exquisitely able to respond to new and unique circumstances, while remaining extremely durable, robust, and a meaningful whole.

• Make forward migration to new and better standards easy. Must be as valuable to leading edge companies as for those who follow.

• Able to be many things to many people at the same time.

• Value producing. Attract voluntary participation because the value of doing so is obvious.

• Exceptionally well-suited to identifying and meeting public purposes.

• Able to protect proprietary property.

No existing institution or system serving labor markets embody all of these characteristics. And while no specific form is stipulated by these descriptions, hierarchical organizations and bureaucracies seem to be especially ill-matched to the challenge, which may explain why so many are struggling to stay relevant. But this dire assessment also creates an enormous market opportunity. The first system that enters this new territory should be explosively successful.

Refining Function:
What Businesses Would Actually Try to Do with O*NET

O*NET is not trying to be all things to all people, even as it is trying to be many things to many people. Its focus is on software tools, databases, and common languages. While these are essential to a wide variety of efforts, they give a distinct locus around which to concentrate attention. Also, since one of the principles is the ability to gain voluntary participation, it is important to try to determine what the key participants themselves are wanting to do rather than what you may want them to do. With this in mind, three functions stand out:

• Create information products and keep them proprietary .

Many businesses consider their strategies and tools for workforce development to be closely guarded secrets. A business may feel that its approach gives it a competitive advantage over its competitors and have no intention of giving that up. This concern is heightened in many so-called "high performance organizations." Any effort to make these secrets public will be strenuously resisted.

• Create a market for their products and sell them.

Many other entities – consultants, educational institutions, and third-party software designers, for example – are in the business of selling the products they develop. Some individual companies are joining this group as well, especially those with a reputation for being pioneers in the field, like Motorola. The products can range from software and databases to intellectual property such as curricula and even strategic frameworks for workforce development.

• Create standards – as widespread or as narrow as necessary.

Standards are necessary for two or more entities to benefit from each other's development work or markets. However, most standards are not universal. Most fill relatively narrow niches so that specialized work can be accelerated. But standards can be appropriate at any scale and, as noted above, there can be great economic value in expanding the use of a standard. Enormous battles are fought over the ascendancy of any particular standard because the economic windfall to the victors can be equally enormous.

Again, these functions predate the DOT and they will be around long after O*NET is a distant memory. These are the functions that a new system either must be better at enabling than the alternatives or must be ready to be relegated to a marginal role in labor markets. Any approach that inhibits these activities or picks sides in a competitive battle will alienate a large number of potential participants.

Refining O*NET's Purpose

With these functions as a backdrop, it is possible to craft a formal purpose statement that is concrete enough to assess organizational alternatives:

Create a system capable of delivering a premium selection of tools and information resources into every labor market in this country.

Each word in a purpose statement such as this has importance and its meaning needs to be clarified and confirmed. It is that part of the process that Dee Hock calls "peeling back the layers of the onion." In this case, the words mean:

…to create…

The system does not yet exist and needs to be built.

…a system…

Right now there are a multitude of efforts but they have not coalesced into something coherent, Something is needed that is a recognizable and meaningful "whole." While this system is expected to have a high-tech backbone, the term "system" means more than just "computers" in this case. It also means the organizational agreements that makes the physical system possible.

…capable of delivering…

There is likely to be a movement of informational tools and resources from place to place, person to person. But the exact form of the movement is left open. The same is true for the creation of the products. A single person could create something, or a consortium of organizations could do so. The transaction could be a one-way sale, or it could be a barter-like exchange. The ideal may even be that the people in the labor market create their own tools rather than having them delivered from the outside.

…a premier…

The system must be the one that people prefer to use, regardless of their other options. It's products should also be able to command a higher price.

…selection…

One size will not fit all. People must have options as to which specific tools will work best for them. Selection means variety. And, more importantly, if no current tool seems acceptable, then people must have the option to create their own withinthe system. It is important that the various choices work together regardless of which pieces are selected.

…tools and informational resources…

Standards, assessment techniques, occupational descriptors, statistical data – all are necessary for a well functioning labor market as well as the tools (e.g., computer applications) to manipulate them in a specific setting.

…every…

The system must be capable of being universal in terms of scope, scale, and access. This means every business, every educational institution, every HR service provider, and every individual worker. No one should be left out. No one should want to be left out.

…labor market…

Narrowly, a labor market is anywhere that there is bidding for human services. The boundaries of such markets can be as small as a single firm, or as large as an entire industry or the economy as a whole. However, since the human "product" is not a static commodity nor is the workplace a static setting – especially in so-called "high performance" companies – both the ongoing development of human capacity (e.g., education) and the planning for new forms of work organization should be considered part of the market.

…in this country.

It is easy to see how this sort of system would quickly cross geographic boundaries. Companies are global, industries are global, and the economy is certainly global. Still, the emphasis is that all labor markets that involve U.S. citizens must be well served. If this system also becomes the world's standard, so much the better.

Refining the Organizational Principles: Borrowing from Visa

Knowing what people are trying to achieve and knowing the general market conditions allows us to posit some initial principles relating to the organizational form that a labor market system would need to take. Those acquainted with the Visa story will recognize some strong parallels. They include:

• No intrinsic advantage of one part of the system over another.

There must be a fundamental equality of participation rights. The system must be equally trusted by people who have diametrically opposed interests, and it must gain the voluntary participation of all. If the system structurally advantages one set of participants over another, then there will be no incentive for the disadvantaged set to continue participation. It must be conducive to new entrants and to the creation of new elements (see below).

• Distributed structure and power.

"Distribution" implies sharing as well as decentralization. Clearly, if anything that can reasonably be done by a more local part of the system, it should be done there, and real decision-making power should follow. But it also needs to be exquisitely easy to associate with others around common tasks without a special or long negotiating process. It should be automatic.

• Growth must be possible without centralized control, yet without risking the system's essential nature or the paths that connect its parts.

One of the powers that must be distributed is the right to create new parts of the system. This is the best way to ensure innovation. Yet, the new part must "inherit" the core features of the whole, however, to ensure continuity.

• Voluntary participation but enforceable agreements.

No one trusts a system that can't enforce its own rules.

• Open and public.

Although this sounds like it is clearly in the interest of DOL, it is really of equal interest to all parties. Keeping a system open allows costs to be spread more widely and provides a constant source of new ideas. More needs to be merrier. Otherwise growth will be too slow to keep up with the changes of the surrounding world.

• Equitable assignment of ownership and cost sharing.

It seems likely that some common assets are going to be created and their availability subject to intense interest. Also, any joint infrastructure that is created will need maintaining. If such items are not equitably handled (e.g., no taxation without representation), then the so-called "free-rider" problem will eventually sink the system.

These principles are not sufficient to create an organization, but it is clear that if they are violated – in any important aspect of the system – then participation will suffer or potentially collapse.

Four Key Elements of a System

Achieving this purpose, of course, is easier said than done. Such a system would be made up of at least four distinct elements, each having to fully embody the principles and characteristics that are described above. The elements include:

1) A general software platform for development and linkage of applications.

The first technical challenge is to devise a means through which the development of software itself meets the principles of no intrinsic advantage, distributive structure, and growth without centralized control. A piece of software must be able to be used as a component with another piece without constantly requiring direct collaboration of the programmers. Otherwise, the software system would never be truly distributive. It must work across a variety of hardware platforms. Every piece must be effectively self contained, but with public standards for how to link with other pieces.

Most software platforms – from Windows to DB2 – are proprietary and would violate several of these tests. Even Microsoft's OLE, which aspires to provide clean, cross-application hooks, would fall well short.

The strongest candidate for this kind of standard is probably COBRA (Common Object Broker Request Architecture) which is both object-oriented and cross-platform. This standard setting process has been organized by the Object Management Group which includes major hardware and software companies including IBM, AT&T, United Technologies and Oracle. The COBRA standards have been in development for more than five years. They are currently in the latter stages of a second round of refinement in which both component interoperability and portability are major emphases. Most interestingly, it is also targeting the development of "business objects" to allow direct modeling of business processes including labor utilization and process design.

If COBRA or a similar existing standard turns out not to be satisfactory, then O*NET would need to launch its own standard setting process.

2) The content and descriptors that apply to any given labor market situation.

As noted earlier, the system would need to be able to allow every firm to construct its own unique language if it so desired. Most likely, however, widespread standards would emerge that would be used by most organizations. Which standards dominate is likely to shift over time, and their place of origination might turn out to be a surprise. The process must be continuously open to the emergence of new standards.

The work that O*NET and the National Skill Standards Board have done, among others, should be more than enough to seed this process. Many individual companies and service providers will want to offer their standards, as well. Although the technical challenge of creating some common ground should not be underestimated. Even constructing some good translations from one standard to the others would be a considerable achievement. Progress is likely to be faster, however, when the focus is on applying them to real business needs rather than the abstract connections favored by some academicians and statisticians.

3) The physical system through which information and tools are transported and organized.

The people with the know-how and the data may be located at any distance from the people with the needs. A physical communication system needs to be able to connect any two people virtually anywhere in the U.S. In some cases, that communication will need to be secure.

The Internet is not mature enough to handle this sort of communication effectively yet, but it could provide an important backbone and is easily extensible. However, some elements may need to travel on a Visa-style data system that is secure.

4) The organization that enables the creation of the system and ensures self-regulation and continuous improvement.

The creation of these first three elements as well as the interaction among a wide variety of independent actors, including fierce competitors, will require a very special kind of organizational framework.

Unlike the first three areas, there is no direct model for the kind of organization that is likely to be needed. It will need to be conceptualized and constructed from scratch, and we have at least one major step to go.

Finding the Organizational Glue

So now it is time for a deep breath. As the description of the eventual system becomes more and more complex, it is clear we have to identify the organizational "glue" that will keep it all together. If nothing remains at the center that reasonably can be done elsewhere, what will be left at the center? And since there are many more than one "center" in this kind of distributed system, is there anything that would be at the center of all of the various centers. You find this, and you have found your glue.

For Visa, the glue turned out to be the notion of the "switch," i.e., the ability to pass a transaction from the bank of the merchant demanding payment to the bank of the cardholder, no matter where those two banks were located in the world. On the one hand, this function was a global one and Visa International could not relinquish its responsibility to ensure that the switch worked universally. On the other hand, any smaller group of banks could devise their own means of moving transactions – i.e., be their own switch – as long as it didn't undermine the switch as a whole.

Moving transactions through a switch was the life-blood, and glue, of Visa. It was a terribly mundane notion, yet wonderfully powerful. Before Visa, there was no world-wide system that could guarantee transactions among banks. Now it's hardly given a second thought.

The notion at the heart of a labor market system might be similarly mundane, and similarly powerful. The need to distribute the development of software requires that there is some way to keep track of the ownership of various pieces so that the original developer can be fairly compensated. Fear of piracy is an enormous deterrent to the creation of more open systems. No one wants to give away their just rewards, nor should they have to.

Tracking ownership of software components – and directing payment back to the owners – is a function whose creation would be an enormous magnet of innovation and effort. It also is a function that must be present in the system as a whole, and in every more local part of the system. It could be the "glue" that is needed.

It is also the kind of function that cannot be embodied in a traditional, hierarchical organization. It needs something new.

Section 3:
The Labor Market Software Registry

With, characteristics, principles, purpose, key elements, and "glue" in hand, it is now possible to be concrete about the shape of a new type of public/private partnership. I have chosen the most direct and mundane name I could think of ("Registry") in order to avoid distractions. Combining all of the prior discussions, we end up with:

Organizational purpose:

• Create a system capable of delivering a premium selection of tools and information resources into every labor market in this country.

Organizational Concept:

• Manage a means to track the use and sale of software components by other software and information integrators so that the producers of those components can be fairly compensated.

Legal form:

• For-profit membership corporation. Non-stock.

Since it is likely to own various assets and be jointly owned by numerous for-profit interests, the easiest and most versatile legal form is that of a for-profit corporation. Borrowing from Visa's experience, avoiding the issuance of stock removes the threat of hostile takeovers and is probably required by the "no intrinsic advantage" principle.

Membership boundary:

• Institutions or individuals who are creating or selling tools or information resources for labor markets.

Since the function of the organization is to register software components, applications and databases, only those that produce such items need to be members. It may make sense to have the users of such products be called members, as well, but if that is done, then their membership rights would be distinctly different.

Since ownership could be registered only to a member, and some products would be the result of multiple individuals or institutions, group memberships would be allowed but would have the same basic rights as individuals.

Likely organizational assets:

• A pool of software primitives and completed applications which can be combined for use and resale under a general license agreement.

One "group member" would be the organization as a whole. There would be a number of imaginable situations where ownership of a product would be shared among the entire membership. Membership would entitle one to these products and would be a major draw to membership.

Software development has often been slowed down by the need to negotiate license arrangements on a case-by-case basis. Many mergers and acquisitions in the software industry have been at least in part due to the fact that its sometimes easier to buy a company than to license a critical technology. Placing and using software components under a general agreement would markedly simplify the sharing of technology.

• Content databases linked to applications

One way to broaden the use of a database is to make it available on-line. This would allow real-time updates without having to reinstall a whole application. Again, some databases would be proprietary and others would be public.

• A means to establish prices for use of components and track inheritance.

The Registry would create an open marketplace for software components. The simplest form would be a price for an unlimited license, but any number of arrangements could be made and should be left to the ingenuity of the members. The only important restriction is that all members face the same price and conditions.

• Registry trademark.

A registry trademark would ensure interoperability of the various software components. Even if software products were technically compliant with Registry standards, they could not cite this fact in marketing materials unless they were registered as members.

Growth dynamics

• Exquisitely low barriers to new members

New members should have the same basic rights and access to Registry property as existing members.

• Devolution of jurisdiction and expansion into new areas

Members have rights to form specialized sub-Registries and take jurisdiction away from a more centralized body as long as they agree to abide by the same basic principles of operation and governance. The same is true for any function that is not currently centralized but where voluntary agreement can be organized.

Benefits to members

• Ease to form a work group with any Registry members to create or market products (and register ownership) or establish new standards (must be Registry wide).

• Right to market any and all Registry products

Since a member can combine several components for sale, there is really no meaningful barrier to market any Registry product. While this is quite different from the way that most software is marketed today, it dramatically democratizes the distribution channels, making it possible for small concerns with good ideas to rapidly capture market share. This will in turn attract the most entrepreneurial minded individuals and companies.

• Testing to ensure Registry interoperability

This is almost certainly an early service of the Registry that will be jointly funded through member dues.

Benefits to customers

• Guaranteed interoperability with any other registry product.

• Right to shift from user license to member license, or back, at any time.

Once a user is familiar with a product, there would be a known economic calculus determining when it is more appealing to shift to a member license (pay dues, but have rights to fine tune your applications with other registry products) or stay as a user. There probably would be an economic incentive on the part of Registry members to keep that barrier as low as possible so that Registry dues could be spread more widely.

• Virtually transparent upgrade path.

Most likely "upgrading" would become a continuous process rather than the discrete process it is now. And the user would have a great deal of control regarding the direction of the refinement, since any Registry member could help him do it.

A Little Crystal Ball Gazing

If an organization like the Registry came into existence, a number of currently difficult things would quickly become possible. Beyond those mentioned in the side bars, just imagine:

• An industry group creating a group member (or group user) to collect data for its own use.

If several firms in an industry were using Registry products, they could easily create a proprietary database to track or identify important trends. Once they had the basic information they needed and were ready to move onto the next thing, they could sell it into the Registry marketplace, and recoup some of their development cost or draw other people into their particular way of doing things.

• An unemployed salesman begins selling software.

A salesmen who was laid off is doing a job search on a Registry product, and is so impressed with it that he decides to create his own business selling Registry products. All he needs is a Registry membership.

• Several federal departments loading statistical data into "free" database.

Rather than investing in huge data processing departments, several agencies decide to publish their data free of charge into the Registry. Multiple parties quickly battle to convert the data into the most usable form and either update their own proprietary Registry databases or simply "point" their applications toward the new data.

• A community college joins with a local business association to create a preemptive training program.

Rather than having to reinvent the wheel, a local college in New Mexico has access to state-of-the-art software components and decides to work with a local Registry member and a group of local businesses to create a life-long learning program for their employees. In-take procedures as well as some curriculum packages are virtually off-the-shelf and can be easily tailored for their purposes. The college gets to focus on what it does best – teach – and leaves the software design and other elements to world-wide experts.

• Trade union creates world-class application only for its members.

An imaginative union was one of the first Registry members and has developed an enormous proprietary library of applications and information that is available only to its members. Its services are so superior that it has gained membership in each of the last five years. Its volume of use alone has made several programmers who supplied some of the union's early software components very wealthy.

• A former manager of a One-Stop Center gains notoriety for being able to translate one data language to another.

Even with all of the standards that have emerged, there is still an art to translate one to another. A public servant in Virginia seems to have developed quite a knack for it, though, and is in constant demand for her expertise. Registry products with her translation components are known to command twice the price of comparable offerings.

While the preceding stories are fanciful, they convey a very different flavor for what could be possible in the market for labor markets tools and resources. DOL would not have to do everything by force of will. It would unleash the energy of a whole marketplace on what is ostensibly O*NET's purpose. DOL could then focus on the always sticky problems of equity and access.

Secondly, DOL would bypass many of the legal and legislative issues commonly associated with privatization. DOL would not have to "give up" ownership, even if it priced its products at virtually zero. The products that have been paid for by taxpayer money could always be made effectively "public" both in access and in cost.

There is nothing keeping DOL from continuing to be an aggressive creator of products and services for the Registry. The difference would be that DOL could recover some of its cost through different pricing strategies; it would face a real market test that would satisfy even the most ardent right-wing economist; and it would benefit from a much wider pool of product developers and innovators.

Section 4: Getting from Here to There
DOL/O*NET Leadership

None of this is to likely to happen spontaneously in the near future. It would take real leadership on the part of DOL and O*NET to further test the idea and then make it a reality. Two paths should be followed simultaneously:

1) Launch a serious effort to construct the new public/private partnership.

• Create a workgroup of a handful of the most creative and publicly-minded individuals and institutions that would likely be active participant/owners in the new system.

• Familiarize them with the concepts expressed in this paper and those surrounding the creation of Visa.

• Have this group work through an initial design process to familiarize themselves with how the process works.

• Steadily expand the circle of participants until there is an adequate critical mass to form the marketplace and work out the initial technical challenges. This process itself can be designed by the workgroup.

2) Begin to conform any policies or RFPs to the principles outlined in this paper and enhanced by the workgroup as it proceeds.

• Use RFPs to encourage experimentation on how these concepts can be applied in the various programs underway in DOL and related agencies.

• Steer any organizational "reengineering" toward more open and market-driven approaches. But a caution: It is probably unwise to try to directly convert any existing program or structure to a form such as those imagined here. Of the more than 20,000 member banks of Visa, none have yet to adopt a comparable structure, even though they have been made infinitely more productive and profitable by its existence.

Conclusion

Creating something like the Registry would clearly signal the start of a new generation of public/private partnerships. As co-owners of a commercial enterprise that creates an expanding marketplace of goods and services, the public and private sectors would control their own destiny. Each would see an explosive increase in their ability to achieve their goals. Each would have played a role in creating a new realm for collective effort and success.

But while it would be a first, it would not likely be a "last." There are numerous other settings where technology is forcing government to reconceive the way it protects the public interest – such as the challenge of governing the so-called Information Superhighway. People's experience with the Registry will lead them to think about how this approach could be applied elsewhere. Each new effort will be progressively easier to organize and to sell. The experience gained in one will cascade to a multitude of others. Eventually, people will be saying, "The Registry? What's so special about that…"

APPENDIX I

Questions to explore.

Working in an iterative pattern from top to bottom.

Descriptions of the various elements, and their corresponding Visa example, can be found on the following page.

APPENDIX II

Some Principles of Open Partnerships

It must be "open" in the sense of being:

• inclusive: The barriers to entry are few, low, and fair. An open setting invites participation, it does not try to restrict it. The door stands open and the threshold is low. More is merrier!

• voluntary: No one should be forced to participate, if the value is not obvious to them. Ideally, there should be so much value being created that people will feel foolish to leave. But the door stands open in both directions, and the threshold is no higher going out than it was coming in.

• open to change: New things can come through the door that may shake things up. The system must give people a means to embrace change when it is required, just as it enables them to fight for stability. Both strategies must be able to exist side by side. Ideally, people adapt to this openness by using it as a pump for innovation and change, making innovation and change its life force.

• able to grow freely to whatever size and scope is possible and necessary to maintain openness: If the walls are fixed, the room will rapidly get crowded. The open door would do no one on the outside any good because they can't squeeze in. The fight for elbow room on the inside would get fierce and ugly. Eventually, the flow of people would be outward rather than inward. But if the walls expand as more people arrive, then the only limit is the number of people that find value in participating and how inventive those people can be in finding new ways to create value that can expand the room even further.

However, it must also be a "partnership" in the sense of being:

• defined: The relationships among the participants must be formally agreed upon. The purpose of the partnership needs also to be defined, including those things that will not be violated in pursuit of goals.

• among fundamental equals: All participants have the same basic rights and responsibilities that cannot be arbitrarily altered or diminished. One participant should not have an intrinsic advantage over another within the organization. However, this doesn't mean that everyone reaps the same rewards, only that the organization can't pick sides unless, of course, a more general principle is at stake.

• bounded: Entry (or exit) is a conscious choice. At any point in time, you can tell who is "inside" and who is "outside" of the organization.

• enforceable: No one trusts a system that can't make and enforce its own rules. Ideally, the agreements should be largely self-enforcing.

• goal oriented: people are participating in order to get something done that couldn't be achieved independently or that can be much more successfully achieved in concert with others.