Monday, August 20, 2007

ABM and Corrections Sentencing

As promised last week, Mark Buchanan at The Social Atom has followed up a post on complex adaptive systems with interrelated agents with a link to a book review in Nature (need to register, didn't do it) on Agent-Based Modeling (ABM). Here's the quote he pulled from the review and the reason why it's potentially earth-shaking for first the academic types, then the practitioners in corrections sentencing:

...these two books are part of an important trend in the social sciences. Both argue for the value of agent-based modelling (ABM) in social science. This approach involves "growing societies from the bottom up", as Epstein has put it, rather than devising analytically airtight theorems from first principles that are tractable but transparently wrong in what they assume and imply about human behaviour.

The aim of ABM is to study whether the macroscopic patterns or regularities that we observe in society, such as price equilibria or the appearance of behavioural norms, can be generated from decentralized, local interactions between collections of agents.... Agent-based models may not describe reality, but they can show how interaction and nonlinearity produce social outcomes that could not be predicted simply by inspecting the behavioural rules.

There has to be some crim-oriented young scholar out there with the interest, brains, and time to pull this together into corr sent simulations and conditions that promote and reduce crime, reentry, or sentencing outcomes. There have been several books that have spent a page or three speculating on this, but none have done it yet. Come on, we're waiting for you.

1 comment:

Gritsforbreakfast said...

Yes we are, indeed, waiting ... git to 'er, young grad student.

However, part of me wonders if the data even exists to supply such a model with the type of multi-dimensional information needed for a true ground-up approach. From my state-level perspective, data aren't even good enough for top-down management (think of the UCR), much less a bottom-up, data-driven model.

We collect certain data for certain reasons, and a new model will need different data for different reasons. I'm just not sure it's out there right now. The first task may be identifying concretely what those new data points are. best,