Creating Policy Innovation in the context of Government Bureaucracies? Could ‘Parallel Learning Structures’ be an answer?

by Colin Talbot

Trying to innovate in a bureaucratic organization, which most governmental and public organizations are, is notoriously hard.

But, despite the many forecasts of its imminent demise bureaucracy is still very much alive and kicking – for very good reasons.

Bureaucracy is Good for Us

Bureaucracy gets an almost universal bad press – but it is actually one of the best inventions humans have ever come up with. Bureaucracy is really good at standardized, replicable, large-scale production of goods and services. In many instances – in both the public and private domains – that is just what is needed.

Apple produce a relatively small range of goods and services that they market around the globe. They want an iPhone 7 sold in New Delhi to be essentially the same as one sold in New York – and it is. How do they do that? Through bureaucracy.

The USAs Internal Revenue Service or Britain’s HMRC collect income taxes. They need to provide a service that is essentially the same in every like-case – they can’t be charging two different people in the same circumstances different amounts of tax. How do they do that? Through bureaucracy.

Bureaucracy solves the problem of complex division and then coordination of labor through a relatively small number of principles: well-defined hierarchical authority; task specialization and roles; functional or task groupings; internal ‘standard operating procedures’. For government or public service we can add the separation of personal and private interests and appointment on merit as being especially important.

Bureaucracy is, in short, good at standardizing and stamping out unwanted deviations. The downside is it also tends to stamp on innovations, making it hard to change.

Policy and Service Innovation

In government there are many areas where bureaucratic standardization is still necessary and no-one has yet come up with a plausible alternative. Tax collection, benefits distribution, licensing, imprisonment, and many other areas require government and public agencies treat people in standardized and therefore fair ways.

Of course, in practice things don’t always work like that and many bureaucracies around the world are still corrupt – but to the degree they deviate from the ideal type they illustrate how important to aspiration to ‘good bureaucracy’ is.

So how can we have the benefits of bureaucracy and innovation at the same time? One of the most interesting attempts to solve this dilemma is “Parallel Learning Structures”, an approach that grew out of the social psychology inspired ‘organizational development’ movement of the 1960s-80s.

The idea of Parallel Learning Structures is set out with admirable clarity and brevity by Gervase Bushe and Rami Shani in their 1991 book with that title.

Let’s start with what Parallel Learning Structures are not. They are not ‘Task Forces’ or ‘Teams’ that mirror the existing power structures and culture of the whole organization – they need to have a different set of rules. They are not ‘Project Teams’ that have a specific plan to implement – they are about learning.

What they are, or should be, are firstly ‘parallel’ in that they reflect (but do not reproduce) the whole system they come from – so different layers and functions. But they should usually avoid including people above or below in the same ‘chain of command’.

They are about ‘learning’ so the internal culture of the group should be egalitarian and recognize everyone has something to contribute. They need to be creative cultures and utilize all their resources. Like ‘brainstorming’ exercises they need to build experimental space for testing out crazy ideas but also develop systems for assessing and judging evidence. They use ‘co-inquiry’ techniques to explore new ideas and evidence.

They do still need to be ‘structures’ – so they need some aims, norms, leaders, facilitators and standard operating procedures – just not the same ones as the entity they come from. They also need clear mandates from the host organization and reporting arrangements.

Once the PLS comes up with proposed innovation or changes they need to do some experimental implementation to test them out in the ‘real world’ of their larger organization. Only after evaluating and further learning from this can the whole organization move on to full-scale implementation.


PLS interventions can be done without any external support, but in most cases I have come across they do have some sort of external input. This can take two main forms: process consultancy (helping to facilitate it) or research and evaluative input. I have done both.

I have used the broad PLS approach many times working within single government and public agencies and even with consortia of agencies (e.g. in one case 8 social services departments). It works.

In these cases the object has usually been solving a policy and service design problem within organizations, but it would almost certainly work across broader systems that involve multiple agencies?

Many of the special units set up to address policy problems never even consider fully their processes or if they are the right sort of teams.

Thus one model often used in Whitehall since the ‘Efficiency Unit’ scrutiny’s of the early 1980s is a small team of, usually ‘Fast Stream’, civil servants. Besides the Efficiency Scrutiny programme they were used for numerous other innovations – the initiation of the ‘Next Steps’ agency creation program; or the formulation of the disastrous Poll Tax policy (as recounted in ‘Failure in British Government: Politics of the Poll Tax’).

These sort of small groups are almost the opposite of PLS interventions – they are closed, not representative of their wider environment, do not draw on expertise, have very success-oriented rather than learning cultures, and so on. As Paul t’Hart has explained they too often descend into ‘group think’, with disastrous consequences. Groupthink, lack of challenge and proper deliberation were also key factors in the twelve ‘Blunders of our Governments’ analyzed by Anthony King and Ivor Crewe

There are other experiences – such as the Strategy Unit in Tony Blair’s New Labour government – that seem to have been closer to the spirit of Parallel Learning but these have been unreported and not properly evaluated.

It will be interesting as the use of ‘Policy Labs’ spreads and we gain more experience to see if the sorts of lessons embedded in Parallel Learning Structures crosses over into policymaking?



One Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.