Two things occurred recently that got me thinking again about the issues raised in Matthew B. Crawford’s book  and the commentary, including the observation made by Robert Pirsig of Zen and motorcycles fame that in “the division of the world into parts … something is always killed” and what is killed is “an awareness and contact with the world before analytic thought has done its (necessarily) reductive work.”

*****

The first thing was that I attended a community forum facilitated by a group of consultants on behalf of the Department of Transport.  The consultants had been hired to review the bus services in metropolitan Melbourne and make them more efficient. 

The consultants were from a well-regarded organisation, a top “tier 2” consultancy (“tier 1” consultancies are the global firms like McKinseys, Accenture, Deloittes).  They were smart and capable, and all deeply engaged with the dream subject matter.  Dream subject matter for a few reasons.  It’s a big brief, and that means big money for the organisation (and bonuses or glory for the individual consultants associated with it).  Even better, it means big recommendations.  On a job like this, the client will want the boldest possible recommendations to show for their money, and this in turn means the consultants will have maximum latitude to investigate, to experiment and to have fun. It’ll be as if they borrowed a cloak of untouchability.  Until close to the end, of course, when it always gets messy.  It’s also dream subject matter because who wouldn’t want to play God with buses?  I didn’t even grow up with a model train set, and I felt the appeal.

So there we were with our Post-Its and our smoked salmon mini-bagels, and straightaway the problem of parts and wholes came up.  For the consultants announced that this was the first phase and in this phase they were looking at routes only; in the second phase they’d looking at timetables, in the third, at access, and so on.  You see the problem, right?  Charged with making the system more efficient – yes, it’s a weasel term in itself but let’s just say it means to transport more people from point A to point B more quickly – they were about to look first at where the buses should run, and only after that to look at when they should run.  As if the routes and the timetables, and the several other parameters identified, were self-sufficient and had no bearing, no intimate bearing, on each other. 

Say if you recommended making a route longer and adding more stops so as to carry more people.  Great!  But then comes phase 2 and you start looking at timetables.  You quickly realise (in fact, you realised at the beginning but had to put it out of your mind so you could continue) that the travel time will be 10 minutes longer.  Using our provisional definition of efficiency, you’re transporting more people but not more quickly.  So you go back to your route and start trying to offset this increase in travel time.  You shave off a stop or two, and recommend that the bus travels on the freeway for part of the route.  Now the travel time is only 2 minutes longer than it was originally, and you’ve only left behind a few potential passengers from the original stops, but because it’s travelling via the freeway the bus now comes into the western part of the city and no longer provides access to all those passengers who work in the eastern part of the city.  And so on. 

There is an eternal regression, or should be, between these variables because they do not affect the efficiency of the bus services in isolation, but in concert.  Not only will the resulting model be deeply flawed as a representation of the reality of the bus services, it will be useless in predicting what will happen when it’s implemented.  With two notable exceptions.  It’s entirely predictable that:

• there’ll be as many unintended consequences as intended ones
• any efficiencies gained will be minor because each manipulation will cancel out others.

*****

In the management consultancy for which I used to work we had a term for the issue of parts and wholes, taken from the handbook for McKinsey consultants called The McKinsey Way.  It’s “mece” (pronounced mee-see) which meant “mutually exclusive, collectively exhaustive.” 

Take the bus review example.  It was defined in mece terms: phase 1: review routes, phase 2: review timetables, phase 3: review access, and so on.  If, however, you were defining it in non-mece terms you might say something like, “The review will make recommendations about:

• increasing the number of people catching the bus
• increasing the frequency of buses
• decreasing travel time
• improving the efficiency of the bus services in Melbourne.”

This is not mece because the fourth point subsumes the preceding three; the points are not mutually exclusive.

In the consultancy for which I worked mece was the highest principle.  Even one’s emails had to be mece.  At no time was it seen as an issue or a conundrum, or an artefact of the act of analysis and representation.

*****

Remember I said there were two things that got me thinking about the issues raised in Matthew B. Crawford’s book?  Well, the second thing was reading Robert Manne’s article in The Monthly about the proceedings of the Bushfire Royal Commission, in particular, the testimony of Serafina Munns, a volunteer working at Kangaroo Ground fire station on that terrible day in February this year.

But more about that anon.