We live in a time when physical and digital infrastructures built in a different economy are wearing out. Over the years, constant budget pressures at all levels of government have made it easy to defer large infrastructure renewal investments. Today, the resulting “rust out” is becoming a critical issue for public sector decision-makers. One place where the pressure to renew infrastructure is particularly acute is in the CIO’s office.
In the case of the federal government, which spends about five billion a year on IT, the Auditor General, in her 2010 Spring Report, identifies a number of large aging systems, the Employment Insurance Program for instance, whose catastrophic collapse would threaten literally millions of vulnerable Canadians. In other jurisdictions aging operational systems prevent direct communications between first responders, exacerbated by the fact that there are fewer people around that know how to run and repair electronic safety and navigation systems built in the 1960s.
This story is repeated at different scales in cities, towns, provinces and territories across Canada.
A conventional response to the need to rebuild IT infrastructure and reduce costs is that of centralized services. The idea is that a single service provider can achieve economies of scale that are not possible otherwise. Unfortunately, our shared experiences have shown that large central service efforts typically don’t work.
How we got here
Our public enterprises are led by sincere, intelligent people, so why then do we find ourselves in a critical situation with regard to information infrastructure? The heart of the problem lies within current governance models and their supporting business culture. The organizational structures for funding and governing horizontal initiatives often do not exist, or if they do, are cumbersome. Combine this structural reality with a culture based on what author David Eaves (eaves.ca) calls the corporate values of a hierarchical system – centralized decision making, a strong delineation between insiders and outsiders, risk aversion, specialization and deference to authority – and the result is a lack of long-term, large scale planning that results in the infrastructure crisis we have today.
Why it matters
If the issue of infrastructure renewal is not on the CIO’s agenda, it will be soon. The risks of not taking action are clearly unacceptable; at a minimum, governments face rising maintenance costs while spending precious money on duplication of services. In a worst case scenario, they could be dealing with the collapse of service delivery and legal proceedings.
The challenge of updating critical IT systems has been compared to refueling an airplane in flight. Essential citizen services must continue to be provided while new solutions are developed and implemented. Not only that, we need to accomplish this in an era of budget deficits and cost cutting. The challenge is to change the way we approach the sharing of infrastructure, to move from empires to collaboration, from ownership to stewardship.
Centralization doesnï¿½t work
Traditionally, when we talk about something like the renewal of internal business services the central agencies position themselves as the obvious choice as developer/operator as they push for a prescribed approach to delivery. This approach attempts to create economies of scale by consolidating internal service delivery, usually on a full cost recovery basis. Line departments are essentially forced to buy services at a fixed price. Unfortunately, this usually antagonizes the “client” departments and agencies who view the appropriate role of the central agencies not as service provider, but as facilitator and coordinator.
In the centralized service model a single organization provides services to all members of the federation. Our shared experiences have shown that large central service efforts typically don’t work. Governments need a new approach, one that moves away from infrastructure empire building toward strategic collaboration between natural partners; beyond the traditional systems ownership model to a framework of infrastructure stewardship of behalf of stakeholders.
To move forward we need to take advantage of natural or existing business partnerships and capitalize on emerging social and open approaches to large scale development.
Strategic collaborative approach
In our suggested model, central agencies become facilitators and enablers, the focus moves from building one monolithic organization, the collapse of which would have catastrophic consequences, toward strategic collaboration for the purpose of building, using and managing infrastructure in a sustainable way. In this role central agencies provide enabling common communications services and ensure that appropriate standards are enforced. They also support the maintenance of an overall architecture to ensure interoperability, security and privacy of information. The bulk of actual internal service delivery is not done by central agencies.
Instead of centralizing everything, we believe that a more sustainable and less risky approach to shared internal services is to take advantage of situations where the clients and provider are already natural business partners. Groups that work together every day across departmental and jurisdictional boundaries have existing relationships and may share, or at least understand, each other’s business needs and culture. In short, there is likely to be some existing level of trust.
For instance, at the turn of the last decade the federal government, as part of its Government on-Line (GOL) initiative, grouped online information in a more coherent manner and began to create business clusters with the objective of presenting a single point of information for services and information.
This clustering approach to the creation of external service points has been followed by a logical evolution toward clustering of internal business services between natural partners, a sort of GOL for the inside. The difference in this round of thinking is that the clustered business partners, not a central agency, provide the service. The idea is to take advantage of economies of scale while avoiding some of the problems of getting too big.
We believe that this approach is more natural in that it allows parties who share business processes and purpose to also share internal services. It also minimizes risk of catastrophic collapse. Another advantage is that professional staff often move up through the ranks within these natural business clusters. On the food safety file, for example, three federal departments/agencies share the management on this important work on a daily basis, Agriculture and Agri-Food Canada, the Canadian Food Inspection Agency and Health Canada. It is not uncommon to find staff with food safety experience moving up through the ranks of government within these three organizations.
This new model supports a more flexible framework that moves away from institutional empire building and infrastructure management toward the development and management of information services between natural business partners working together where the operating partner becomes the steward of the system for their partners and not the owner.
We know that empires don’t disappear overnight and collaboration takes trust, which is why it makes sense to leverage existing natural business clusters to implement this sustainable infrastructure model.
An information ecosystem for government?
If we extend this idea of natural business clustering across jurisdictions, there is an even greater opportunity for infrastructure renewal, cost savings and service improvement with the vertical integration of internal information services between levels of government. This cannot happen everywhere at once. But where two or three levels of government have a clear business requirement to share the information being maintained at each of the different levels, this new approach makes sense.
There are lessons that can be applied to the development, maintenance and governance of these clusters from the success of distributed open source software development. This extended model evolves over time and is based on shared standards and open architectures similar to the emerging Open 311 amongst municipalities.
Early examples of this kind of shared service delivery targeting business is the multi-jurisdictional BizPal, which ties together information from federal, provincial and municipal levels of government to make the process of starting and operating a business simpler. Another example of this type of approach is the Health Information Access Layer (HIAL), an architecture and set of standards that stretches across all jurisdictions.
With the advent of social networking tools, the maturation of open source development approaches, not to mention cloud computing, there are new opportunities to build an information infrastructure across government that will be diverse, coherent and sustainable. Taking advantage of these opportunities will require us to stretch our ideas of governance and funding while adopting an ethos of collaboration and sharing. Within governments, collaboration platforms like the federal GCpedia or the Ontario OPSpedia are beginning to demonstrate new ideas of distributed governance and funding, while providing a forum for the “coalitions of the willing” to come together.
Being better with less
Empires and their hierarchies don’t disappear overnight and building collaborative relationships takes time, but if governments can get over the cultural issues associated with empire building and ownership and move toward to a business model that encourages collaboration and stewardship, then a natural outcome will be a service model that allows affordable and safe infrastructure renewal, supports and rewards horizontal and vertical integration of systems, and the improvement of associated business processes.
In other words, governments need to take a more strategic approach to developing and sustaining their information infrastructure. Rather than trying to maintain a centralized model that is high risk and has proven to be unworkable, they need to encourage line or business organizations to take on a new role of service custodian within their natural business clusters. These organizations already work in partnership both horizontally and vertically, so a change in the information infrastructure governance is a natural progress that maintains these systems close to their users.
In many cases, new governance and funding will require changes in legislation to allow the collaborative infrastructure to be developed. For example, in the Government of Canada most line departments are prohibited from funding and operating systems that support client activities outside their mandate. If we can evolve our central agencies, support collaboration and adopt standards along with a shared distributed architecture, then maybe we can not only reduce costs, but also get a higher return on our investment and develop additional systems beyond our current budget constraints. In other words, ensure future infrastructure capacity and be better with less.
What you can do as a leader
● Look for the big picture; exploit the power of “we.”
● Work to create or evolve collaborative funding mechanisms for capital and evergreen budgets.
● Build governance mechanisms that include business, service delivery and policy aspects that undertake long-term investment planning.
● Change the rules. There may be legislative barriers to common sense; work to change this. In the meantime, coalitions of the willing can do a lot.
● Organize internal service delivery around natural business clusters.
● Encourage the adoption of open common architecture and standards (these may be emergent rather than fixed).
● Use a portfolio approach to manage IM/IT investments (recommendation from OAG).
● Do whatever you can to mandate sharing of infrastructure; make it hard to “go it alone”.
Thom Kearney and Christopher McBean founded Rowawnood consulting in 2003. Kearney recently completed a three-year executive interchange with Treasury Board Secretariat where he led a project to review internal service renewal projects. McBean has worked for over 30 years in S&T (www.rowanwood.ca or firstname.lastname@example.org).