A computer program can loosely be defined as as set of rules that turns an input into an output. This set of rules is fixed at any moment in time, yet it evolves with time. Law is pretty much the same. It's all about taking an input (say, a crime) and applying a set of transformations to it in order to generate a specific output:
#set($presumedkiller = 'John Doe')
#if($presumedkiller.isGuilty() == 'true')
$presumedkiller.setProperty('home', 'jail')
#end
In the above case the "software" took a man, assessed its status and defined what should be made of him. Exactly like what law does. The law evolves to match a country's situation and achieve specific outcomes. Laws have aims. They are meant to make things happen the same way software is.
The main difference between software and law is that judging requires human assessment. Defining what's the input -did the killer intended his kill?- requires an assessment. However some law fields are more technical than others and leave less space for interpretation than crime law does.
If laws are pieces of software, then lawyers are hackers. I'd like to focus on the example of tax lawyers. Their job is to study a given piece of legislation and turn it around in order to lower the amount of tax money their clients are to pay. Anytime a governement passes a new law meant to close legislative loopholes, tax lawyers study the dispositions and find ways for their clients to benefit from them. This is very similar to the behavior of hackers who scan new software releases for fails they could exploit to their benefit. The government plays the same game of cat-and-mouse with tax lawyers that software providers play with hackers. The same mindset is at work in both cases.
An interesting byproduct of this observation is that I think there is a market for a piece of software that would modelize the effects of a piece of legislation on a person's wealth based on her financial situation. Sure, the original piece of software reflecting the state of law in a given state would take a long time to write and keeping it up to date would be hard. Yet, since many people who can't afford the services of tax lawyers would end up paying less taxes thanks to such a piece of software, an opportunity exists.
Such a piece of software could get built as a web service by banks or accounting firms in order to better serve the needs of their customers. The market opportunity is interesting. Mint.com is already very close to providing this service but from what I understand they don't offer potential savings related to legislation changes - yet. One could argue that Mint already does it since banks create new offers when an opportunity arises from a new law and Mint will suggest those offers to its users, but a piece of software could still help banks find new opportunities to reduce the amount of tax paid by their customers when new laws are passed. Maybe some banks have written such a program already.
On a related note, I think we could greatly improve the outcome of any given law using computer modelization to predict the effects the law would have on its target population. For instance by using a simple software program and defining a representative population of taxpayers I'm pretty sure one could be able to predict the outcome of any given piece of tax-evasion related legislation. The number of people likely to flee to a neighboring foreign country with lower taxes could be estimated based on how the law is phrased. This might reduce the number of useless or counterproductive laws. Could be useful.
Here's my guess: computer modelization of human activity is going to take a greater and greater place in lawmaking in years to come. While trying to use software to predict the actions of any given individual will undoubtly prove very tricky, the same set of methods applied to statistically significant samples of the population will yield interesting results. However modelization will not replace policymaking since the choice of what to modelize and how to do it will in itself have political implications.
I'd be very interested in finding out more about whether and how the effect of laws being discussed by the French Parliament gets modelized and estimated during the lawmaking process. How long before the Nobel Prize of Economics goes to a computer scientist?
