divider image

Organization design blog

« Change, transformation, tools, levers and systems  |   Main   |  A busman's holiday »

Accountability: is it a design concern?

06/12/17  2:42 PM 

We've had several discussions this week on 'single point accountability'. This sounds straightforward as a concept. Like financial accounting, 'accountability is about giving a reckoning of the actions taken-—and the actions not taken-—that led to the final outcome. Just like in accounting, where your balance sheet must add up correctly, there also has to be a balance in performance accountability'.

Unfortunately, mostly, we think of accountability in relation to assignment of blame (loss) rather than in relation to credit (profit). If you're accountable, you take the blame for what goes wrong. CEOs are usually held accountable for wrong-doing that occurs in their organization and in many cases resign as a result. Sir John Rose, formerly CEO Rolls-Royce, and Martin Winterkorn, formerly CEO Volkswagen are two cases in point.

While they were still at work, both Sir John Rose and Mr Winterkorn received accolades for their leadership and credit for rising sales and share prices. Now, they face pressure to explain why they should not be held accountable for the bad things that happened on their watch as well.

Resignation doesn't solve the issue of what causes something to go wrong: why has the loss occurred? It's often hard to find out. Faster, Higher, Farther: The Volkswagen Scandal by Jack Ewing, a new book on the Volkswagen emissions problem attempts to explain how that situation arose.

But, as in similar situations, he can't give a clear-cut answer. He cites a range of factors involved – tolerance for breaking the rules, lack of checks and balances, employees who had reservations about illegal software had no place to turn, and the personality of Ferdinand Piech – a demanding and fearsome boss.

Finding a single accountable person, close to the work process or issue involved, In this sort of situation is probably impossible, so the CEO takes the 'buck stops here' position.

Many organisational redesigns involve doing a RACI exercise (putting names against each of the four aspects of accountable, responsible, consulted, and informed with an emphasis that there should be only one person accountable) in order to mitigate the risk of diffused accountability. That is fine in some situations - for example a single process operating within obvious organisational boundaries. In most cases it is insufficient. There are few work processes which are not complex. Many cross organizational boundaries.

Take the fashion industry's supply chain which the Bangladesh Rana Plaza catastrophe highlighted. A 2013 Panorama programme, "Dying for a Bargain", review commented that '[It] again showed the total fallacy of factory audits as finish times for workers were falsified and workers locked into factories. Research also shows how many western fashion buyers base their calculations on misleading industry standards set on 100% efficient factories.'

In the networked situation, there are still people responsible for doing work for which they are held accountable by someone else but it doesn't work neatly in a hierarchy to the top single point of accountability. It's not like the nested Russian Matryoshka dolls.

Additionally. accountability is not only about the people: individuals and their relationships, but also about the processes, systems, checks, and balances that enable people to take accountability or to be held accountable. And, in my experience, these formal system aspects are rarely examined as part of doing a RACI analysis or other work to assign accountability.

Assigning accountability in networked organisations is hard. The Oxford Handbook of Public Accountability notes that: 'In all networks, responsibility is shared, ruling out any single point of accountability and giving rise to intensified problems of "many hands". If no one person or body is formally in charge, nobody can be called to account for the network's collective outcomes or made to impose remedies in the wake of acknowledged failure'.

Our discussion on accountability was in relation to a work process that crosses several organizational boundaries – both internal and external. How do we make it work more effectively in the absence of any obvious single accountable owner, and the reality of assigning one?

Peter Bregman in an HBR article, The Right Way to Hold People Accountable, suggests five ways for ensuring people take accountability. He is focused on individuals but his suggestions are usefully extended into organisational policies, practices, systems and processes, within and across boundaries.

Clear expectations: along the work process which organisation is delivering what and accountable for it? Can this be stated and reinforced through contracts, service level agreements, or similarly formal arrangements?

Clear capability: do all the organisations in the work process have the right capability – technology, time, skills, money, data - to deliver what they are expected to deliver? If not, how can others in the network support or help reduce any gaps?

Clear measurement: is the work process being measured across the whole process and not just in discrete steps? How often are the measures reviewed? How transparent is the whole process to every organisation along it? Can each party clearly see at all times their contribution to the overall outcome?

Clear feedback: How quickly are the feedback loops operating in the case of failure? Does everyone have equal rights in highlighting and starting to solve any issues? Are the channels available to ensure rapid feedback? Are decision rights balanced to ensure the work process can work effectively all the time?

Clear consequences: What are the agreed consequences for failure in any part of the whole process? Are the consequences equitable along the process or is one part likely to be blamed more than another in the event of failure. Is there a better way to operate that avoids blame rather looks to collaborate and problem solve rather than to finger point?

There's much research in the field of networked governance that shows both the high value of good relationship building within the network and the formal ties of bridging and brokerage through systems and processes.

The ultimate aim is for everyone to feel accountable for the success of the overall process and be supported by systems, processes and policies which support this. In my view accountability is a design issue, perhaps more than a people/relationship issue? What's your view? Let me know.

Post a comment




back

  • Copyright 2017
    Naomi Stanford
  • Naomi Stanford is an author, teacher,
    consultant and expert in organization design.
  • You can contact Naomi at naomi@stanford.cc
    or our contact page here.
  • Website design & CMS:
    SiteStrux, Inc.