This is a post about Spatial Thinking in the context of the application of justice in New Zealand.  Only a few pretty pictures to go with this one I’m afraid.


Growing up in relatively small-town America, I always knew to be a bit careful when driving through the wop wops – one could never be quite sure of the quality of mercy the further you got from civilization.  The picture above is from To Kill a Mockingbird, a classic tale of small town American-style justice (and a very influential book and film in your blogger’s social development); you could never assume that every small town had a resident Atticus Finch to come to your rescue.  I think we’d all like to assume that justice is applied evenly and consistently across our fair country, but in an interesting story on National Radio the other morning, a researcher reported on how sentencing rates vary depending on where you’re being sentenced, particularly between metropolitan and provincial areas.  What he found was that if you’re arrested for, say burglary, you’re 3 times more likely to  be incarcerated in the provinces than if you’re in a metropolitan area.  For recidivist drink driving, it’s more like 12 times more likely!  His paper hasn’t appeared in the Australian and New Zealand Journal of Criminology yet so it’s a bit hard to get a lot of detail but clearly there’s something going here, and there would appear to be strong a spatial component.

What piqued my interest was that this is an example of what is known as Spatial Thinking, which is really nothing more than trying to take into account how location may influence a phenomenon – is where something’s happening an important part of why or how it’s happening?  And which elements of the location are important?  This particular analysis is being applied to something that’s not often thought of as spatial (unless we’re comparing countries perhaps) and which might surprise a lot of people.  So it would appear that for this phenomenon, the location IS important.  And you could probably envision a map showing how these rates vary from place to place.

But while location is clearly important here, is there something intrinsic about the location that causes this to happen?  Maybe it’s something in the water, but I suspect not.  This strikes me as a example where there’s clearly a spatial component, but it may not really be anything about the specific location that is a causal factor.  It’s not like liquefaction, say, where Peter Almond would probably tell us that it happened where it did because of the nature of the sediments that existed there pre-earthquake, like former river beds or wetlands.  I suspect that what drives the spatial side of this is more a reflection of the social/human environment that has established itself there.  According to Mr Goodall, there are no sentencing guidelines for crimes such as this (though there are for other types) so judges are left to apply their own (for lack of a better word) judgements.  There may be some convergence among judges in an area if they discuss their views and come to some unconscious consensus.  I wonder too if it’s also a reflection of judges being far less anonymous in smaller, provincial towns, and so are more responsive to public opinion.  That’s pure speculation on my part as I know nothing about how one should form sentencing opinions.

I’m reminded too of some of Simon Lambert’s work on post-earthquake resilience in Maori communities.  He’s seen that there’s a lot of variation in how people responded, and while there might be some spatial pattern to it, it might be more influenced by non-spatial things, like income or degree of integration into a neighbourhood; things that very person to person, but not necessarily from place to place.  You can map it, and perhaps infer a pattern, but the driving force behind the phenomenon is likely not spatial in nature.  And I suspect that’s the case with the sentencing rates.  Clearly there are differences, but the actual locations (and the characteristics of those locations) may not really help to understand the process in any great depth.  It might be tempting to calculate something like the number of incarcerations per hectare, but that’s ultimately a pretty meaningless number.

What makes this even harder to fully tease out is that social attitudes can clearly vary from place to place, and it can be a combination of people and place.  I look at where I live, in Lyttelton, and it’s had an increasingly Bohemian reputation for a while now (quite different from even ten years ago), with an eclectic assortment of artists, musos, and oddballs.  Why?  Well part of the answer is place – the views, the physical isolation, the gorgeous landscape (yes, I’m biased) but it also has to do with a (pre-earthquake) collection of run down buildings with very cheap rents (and proximity to the Wunderbar.)  The socio-economic differences between, say, Aranui and Fendalton perhaps stand out on several levels, but how much does it have to do with the specific place, per se, and how much does it have to do with the demographics of the people who live there?  They are intertwined to an extent – just how much and in what ways is beyond my powers of reckoning.  My point here is that location is often an important component of understanding a problem, but not in every case.

Another interesting aspect of this work is that Mr Goodall points out that the districts that judges preside in are not the same size.  There’s a classic problem in geospatial analysis, the Modifiable Areal Unit Problem (MAUP), that this touches on.  The gist of it is that you might get different analysis results if you aggregate areas using different boundaries and, in this case, may make it more difficult to interpret the results.  I haven’t seen a map of these districts but it’s quite likely that the provincial districts cover much larger areas than metropolitan areas, and likely have different populations (and demographics).  The MAUP will likely pop up as another post on another day.

Spatial thinking is, I guess, just a different perspective, another way of trying to understand a phenomenon.  With a lot of natural resource problems, it can be a great aid as it might help explain why certain things happen in certain places.  It won’t always help to shed light on the problem, but as I tell my students, it’s another arrow in your quiver of ways to understand the world.

Think about that next time you’re driving through Culverden and thinking of knocking over the dairy.  (And thanks to Peter and Simon for allowing me to refer to their work.)