There’s a class of crimes that don’t have immediate, identifiable victims. Clearing native vegetation, smoking cannabis, hosing a driveway during Level Three water restrictions, parking in a loading zone.

On the weekend I saw a sign at Minnamurra Rainforest walk that said “No swimming or diving. Penalties apply.”

Last month the NSW Parliament made the act of compensating the mother of a surrogate baby punishable by two years imprisonment.

Is criminalisation an effective way to control behaviours that are deemed to cause social or environmental harms?

The oft-quoted success story is Mandatory Seat Belt laws. So I thought I’d go back and look at how these laws achieved such a momentous change in behaviour.

I wasn’t prepared for the story that emerged. In short, seat belt laws, combined with continuous enforcement and education, HAVE modified the behaviour of all but a small percentage of the population. Nowadays, practically everyone belts up. But what seat belt laws HAVE NOT DONE is make us a whole lot safer. Amazingly, after 40 years, repeated analyses by credible statisticians show that we are little safer than before the laws were enacted.

Subtle forms of social resistance appear to have almost completely negated the hoped-for impact of these laws on road safety.

I didn’t think that would happen. Neither did the authorities. The story has important implications for anyone attempting to control behaviour through criminalisation.

Here’s the story, and below it are some lessons.

What lessons can be drawn from the history of seat belt laws?

1) If you can avoid using criminalisation as a tactic, do so. Criminalisation causes resistance and unexpected blow-backs can fundamentally undermine your efforts.

2) Ensure you have a believable, science-based, clearly articulated case, supported by respected, independent voices.

3) Don’t consider criminalisation until you have overwhelming support and a high degree of voluntary compliance in the target population. In effect, criminalisation should only be considered when it reinforces exiting social norms. Criminalisation may therefore be useful at the end of a long process of voluntary change to control the behaviour of a very small number of chronic resistors.

4) There will always be a small number of resistors and they are likely to cause a disproportionate amount of harm. Therefore understand that criminalisation requires a significant, endless investment in monitoring and enforcement. Weakness in this area is likely to fundamentally undermine the impact of legislation.