REPLY Policy paradox and SD (SD6699)

[SDMAIL Bill Harris]

Posted by  Bill Harris <bill_harris at facilitatedsystems.com>

"SDMAIL Bill Braun" <bbraun at hlthsys.com> writes:
> > I wonder if our concept of accountability would benefit from a mental 
> > model change? Accountability is commonly taken to be something expected 
> > by others (by virtue of being imposed or implied). What about 
> > accountability that is chosen for oneself?

Bill B.,

Several ideas come to mind from your note, which may be a bit of a
diversion from your original idea but which may be useful.

  Is what you're describing better called (personal) responsibility? 

  One of the benefits an OD approach brings the world, similarly to that
  of SD, is helping us talk at a lower level on the so-called ladder of
  abstraction -- in other words, helping us be more concrete in our
  descriptions.  I wonder if accountability and responsibility are some
  of those words high enough up on that ladder so that it's not really
  clear to us what they mean -- or, rather, that it's clear to all of us
  but we each have different understandings.  I wonder sometimes if a
  moratorium on using the word "accountability" would help.

  Sometimes I think accountability is used synonomously with punishment
  or the threat of punishment.  Punishment, of course, is a ways still
  up that ladder, as it has a wide range of potential implementations,
  ranging from (in a corporate setting, for example) your boss saying
  "Don't do that again" with a smile all the way to getting fired,
  getting fired and getting sued, or getting fired and getting charged
  with a crime.  If I were defining this (externally-imposed
  accountability) in an SD model, I might call it "management pressure."
  In the public sector, accountability often seems to mean the threat of
  getting fired or not getting re-elected.

  Thinking of the range of levels of management pressure brings up the
  idea I want to toss out.  Years ago, I did a model of an expense
  management problem (http://facilitatedsystems.com/expmgmnt.pdf and its
  references describe the project and a related model).  I realized that
  there was one parameter in the model I didn't know: how much a manager
  would change her spending habits in the next period, based on spending
  variance in this period.  I decided to make that a free parameter, I
  decided it represented "management pressure" (you might call it
  "accountability"), and I decided to analyze the sensitivity of the
  policy recommendations to that parameter.

    In good SD practice, I should ask you to write down now what you
    think I discovered, for some may find it tempting later to say the
    result is obvious, and it would be interesting to see if it is.  So
    write down your prediction!  :-) 
    
    You've written it down, right?  :-) 

    I tested two basic policies (plus some variations): the existing
    one, which created the problem, and a revised policy, which worked
    (both in the model and in real life).  

    I discovered that management pressure didn't have much impact at all
    on the behavior engendered by the revised policy.  

    Management pressure had significant impact on the behavior
    engendered by the existing policy.  If the simulated management
    pressure were high, then spending variance was initially quite
    improved, but then it got much worse.  Overall, increased management
    pressure (accountability) made things noticeably worse.  (At this
    point, I imagine the "that was obvious" group is thinking of root
    locus plots.)

    Reduced management pressure (accountability) under the existing
    policy actually made things somewhat better.

    Out of that, I devised a hypothesis that has served me well:

      A manager's primary responsibility is to craft an organizational
      system that achieves the (good) goals of the organization
      (http://facilitatedsystems.com/weblog/2007/11/managers-job.html). 
      
      If a manager succeeds at that task, it may not matter too much if
      that manager is a hard or gentle taskmaster, as long as she
      doesn't do things that disrupt the workings of the organization's
      social system.  

      Because managers are fallible, they can't guarantee they always
      succeed in their primary task.  Thus a gentler approach (reduced
      management pressure) may be a wise and risk-reducing approach, for
      it won't hurt performance if you get the system right, and it may
      help performance (relatively) if you get the system wrong.

    Incidentally, that model also has implications for why using a
    credit card may be riskier than using cash for personal purchases.
    See the articles for more.

> > Those who show up will be ready to work. They might talk about others 
> > who ought to be there to help, and ask, what invitation would bring 
> > them into the group?

Back to the central focus of your note, take a look at the precepts of
Open Space, one of which is something to the effect of "Whoever comes
are the right people" (e.g.,
http://www.awcnet.org/awcconf06/handoutspresentations/openspace1.pdf).

Bill
- -- 
Bill Harris
Facilitated Systems  
Posted by  Bill Harris <bill_harris at facilitatedsystems.com>
posting date  Sat, 15 Dec 2007 09:54:22 -0800