REPLY SD Impact on National Government policies (SD6196)

[System Dynamics Mailing List]

Posted by  Tom Forest <tforest at prometheal.com>

Jim Hines related an interesting anecdote about growing like weeds. 
Ecologically, weeds are fast-growing plants that take advantage of 
disturbances like fires and floods to propagate widely, inadvertently 
creating a more hospitable environment for slower growing but longer-lived 
species. In a weedy environment, a few species predominate. There is much 
more diversity in the longer-lived species. In a forest, though, only a 
few will be large and pervasive like canopy trees. I sometimes wonder why 
SD continues to exist at all, as obscure and little-known as it is. But 
while it is clearly not a weed, it is just as clearly not a canopy tree. 

Let me take up Chander Chawla's challenge, gleaning information for my 
analysis from the discussion to date. 

-- Problem statement: the perceived level of SD use for governmental 
decisions is far below the level desired by SD practitioners. 
-- Reference mode: slow growth to a low level of use--perhaps still slowly 
growing (in absolute numbers), perhaps plateaued or even declining (in 
relative terms). 

Several people have discussed their perceptions of the level of use. The 
questions seems to be whether it is significantly higher than what we see, 
and if so how could we improve our knowledge. As a volunteer consultant on 
this project, I recommend that we defer addressing this question to Phase 
II of the project. But let me point out that the membership of the society 
has two different membership modes. One is the relatively long-term 
members with primarily academic and methodological interest. The other 
mode is relatively short-term members whose expertise is elsewhere but 
wish to glean enough about SD methods to help them in their area of 
expertise. Hence there is likely to be a much larger number of lapsed 
members (whose primary knowledge domain is not SD) who no longer have 
regular communications with the society than there are existing members. 
Can someone produce an estimate of how many individuals have ever been 
members of the society? Of course, some of them will no longer be living, 
but it will give us an upper bound. 

A generic market penetration model would be a good starting point. What 
are the competing methodologies? What factors affect SD's relative 
attractiveness? What can we in the SD community do to enhance its 
attractiveness? How long is the sales cycle? What level of effort is 
required? What are the latencies in bringing that effort to bear? The key 
leverage point will be this: how should our effort as a society be 
allocated between selling directly, adding more sales staff, or improving 
SD's attractiveness, or some combination thereof? Put another way, how 
should we budget between sales, recruiting/training, and product 
development? The classic SD answer is that all need to be done in a 
coordinated way or there will be a bottleneck generating oscillatory 
boom-bust behavior with periods depending on the time constants of the 
three requirements for growth but likely to be some product of them. 

The next step I would take in a real consulting engagement would be to 
build a consensus on what the competition is, what the salient attributes 
of attractiveness are, how quickly and by what means they could be 
changed, etc... for all the levels and rates. Then I would run some 
experiments with different allocation choices among the three promotional 
activities. Of course, since SDS is a voluntary association of dues-paying 
members, actually implementing any recommendations we came up with would 
be yet another challenge. We all do whichever of those three activities we 
have interest in and availability for. Might I propose a Governmental SIG 
for those interested? Might I also propose an active policy to slow the 
turnover of members whose primary knowledge domain is not SD, especially 
(for the purpose of this problem) governmental employees? 

In Phase II we could address the limitations of our perceptions and try to 
generate policy recommendations for enhancing those perceptions.  If we 
were successful in implementing those policies, it might turn out that our 
perceived level of use meets or exceeds our desired level. In that case 
the recommendations from Phase I need not be implemented. My professional 
opinion, however, is that even with perfect knowledge we would find that 
the actual level is well below the desired level so that after all this 
work we'd still have to implement our Phase I recommendations. 

Finally, as many on this list have said over the years, as long as we sell 
SD we fail. When we sell solutions to others' problems we succeed. 

Tom Lum Forest 
Forest Grove, Oregon 
Posted by  Tom Forest <tforest at prometheal.com>
posting date  Wed, 17 Jan 2007 23:34:17 -0800