231 – A New City O/S – Stephen Goldsmith and Neil Kleiman.

875 words (4 minutes reading time)                                                               by Lancing Farrell

distributed network

I have also been reading Goldsmith and Klein’s bookA New City O/S. It is a very interesting treatise on a potential future model for local government, particularly the argument for distributed governance. As Colin Weatherby has described, it is a researched and expert work.

I found the concept of distributed governance quite interesting. In some ways, all councils already operate using a version of this model for some services. Typically, this would be in community services where the council, State and not-for-profit organisations often combine to provide a facility, grant funding, and the actual delivery of the service. I think it would be a big challenge for many councils to adopt distributed governance across all services because of the loss of control.

The many sources of accountability for councils tends to make them risk averse and control focused. This is also a key argument for change made by Goldsmith and Kleiman.

“… the very rules designed to frustrate graft and waste now also frustrate employees who are ‘continuously monitored and investigated by auditors, judges, budget examiners, performance evaluators, legislative committees, public watchdog groups, clientele associations, citizen bodies and media organisations eager for a good scandal’.”

The limitations on delivering services that meet the needs and expectations of citizens are self-imposed by the sector and its regulators. The sources of local government accountability has been the topic of a previous post. In Australia, as in the US, it is a constraint on innovation and improvement.

It is worth re-capping on what Goldsmith and Kleiman describe as distributed governance. It is hard to find one succinct definition of what it is, and there are many partial descriptions where there is discussion of why it is needed and how it works. I have assembled several of those descriptions in my attempt at a definition.

Distributed governance places the city at the hub of civic work, with ‘real communications, coordination and connections’ across a broad range of third parties, including residents, contractors, community organisations, local institutions and non-profit and for-profit organisations. It leaves behind strict rules and tight control of information. It is customer-focused and has speed and flexibility to find partners to provide services with autonomy. At the centre, the city provides standards and guides overall system operations with clear directives. It is less involved in delivering services, and more involved in ‘setting roles and rules for conduct, quality, equity and privacy’.

It is clearly intended to be a distributed system. In brief there are three main types of systems or networks identified by Paul Baran, a pioneer in communication networks and the development of the internet.

Baran networks

Most councils use a centralised model in both how they work as an organisation and in the way they deliver services to their community. There might be a decentralised model for some services, typically internal business partners or partnerships for community services. The flexibility and responsiveness potentially available from a network is certainly attractive when considered in comparison with the rigid bureaucracy common to most councils. However, does it adequately consider the political nature of governance in councils?

For it to succeed in local government, a distributed model would need a clear and common purpose. What direction is it going in and who is steering? Without this clarity it could easily fragment into a set of networks, much like the functional silos that exist in many councils. It would also need information symmetry. This is a key point made by Goldsmith and Kleiman, and why they have focused so heavily on the need for a new O/S to enable distributed governance. The open and free flow of information between all parties is the way common purpose is created and fulfilled. But would this be enough?

Working in a distributed network, in some ways akin to network alliances more common in business (although I am sure Goldsmith and Kleiman would say not), is a departure from the way most councils have traditionally operated. Councils rely on contracts and other formal agreements with other entities when working together. It gives certainty and control. Politicians are often uncomfortable leading organisations responsible for ensuring that a community receives services if those services are being left to others to deliver as and how they can. It comes back to the accountability question – who is responsible for the services delivered using distributed governance?

My final comment in relation to A New City O/S is in relation to measurement. Goldsmith and Kleiman emphasise the importance of measures (and using technology to make and report on measurements) in ensuring that the O/S for distributed governance works. The need to ‘migrate from the measurement of activities to the measurement of outcomes’ is mentioned repeatedly. As is the importance of measuring problems solved, rather than compliance with rules. This I strongly agree with.

Measures give control. If they are effective measures of fulfilment of purpose as described by John Seddon, or, in the case of Goldsmith and Kleiman, the outcomes or problems solved, and they are used by staff delivering the service to understand performance and make corrections as required, they can immediately improve services.

Maybe, of all the ideas in A New City O/S, changing the way we measure is the most relevant and doable.

81 – Designing a performance management system. Part 2 – A process.

Posted by Lancing Farrell                                                              560 words

measures process a

This is the second post in this series. The first post discussed the type of measures needed to manage for performance. This post looks at how you can design an integrated performance management system.

The starting point for developing a performance management system is the ‘performance logic’ of the organisation. What are the most important indicators or measures of performance for your organisation? What are the key requirements of customers and the main strategic business needs of the organisation? Continue reading

67 – Design a management job for high performance. Part 1 – some theory.

Posted by Lancing Farrell                                                              1250 words


In the previous post I discussed a tool that you can use to test your current job design to see whether it has been designed for high performance. In this post I will elaborate on the theory behind the tool. This is a long post but I didn’t want to split the story. Sorry. In the next post I will attempt to apply the theory to the design of three local government management roles that I am familiar with.

Simons’ starting point in discussing the design of high performance jobs is failure to implement strategy. Why is it that organisations with clear strategy, access to resources and developed relationships still fail? He points out managers being too complacent and slow to respond, instead of being entrepreneurial. Problems coordinating activities across functions. Decision making is fragmented. Costs are excessive and eroding surpluses. When these symptoms become evident senior managers start to wonder whether they have put the wrong people in critical jobs. However, Simons says that the problem is systemic across the organisation. Continue reading