• Dr Sarah Cowie in the School of Psychology at the University of Auckland.

While Dr Gilbert makes a relevant point about the relation between one's income and the "affordability" of a fine, his argument of "fairness" is seriously undermined by the fact that there is very little evidence that traffic fines actually stop anyone from committing driving offences.

For example, the NZ Police's road policing offence data doesn't show any systematic decrease in speed-camera fines issued between 2009 and 2016.

If the fines issued aren't decreasing, presumably the behaviour that we're trying to punish is also not decreasing.


Perhaps this lack of change is because most of the people who receive speeding fines are in the higher income bracket where fines are small annoyances and not big deterrents.

Alternatively, the problem isn't related to the magnitude of the punishment – regardless of income, fines don't seem to stop us from speeding.

Indeed, Dr Gilbert cites public approval of the Finnish system of income-related fines, but he doesn't mention anything about whether such fines have actually changed Finnish drivers' behaviour.

Where is the fairness in imposing penalties that have no demonstrated effectiveness in reducing bad behaviour?

It's hard to measure the impact of fines on behaviour; the frequency of fines issued will depend not only on how likely drivers are to break traffic laws, but also on the number of speed cameras and checkpoints set up, and the number of drivers on the road.

Additionally, this figure doesn't capture every single speeder; some drivers are simply good at not getting caught.

We can't say definitively that the lack of change in fines issued reflects an absence of change in the probability of speeding, but there doesn't seem to be much evidence that fines are making a meaningful change.

One way we can assess how likely something is to be effective is to look to science.

Unfortunately, science also doesn't support fines being effective ways of reducing unwanted behaviour.

For example, we know that the greater the time between our actions and their consequences, and the lower the probability of those consequences actually occurring, the smaller the impact of those consequences on behaviour.

This is why it's so hard to stick to those New Year's resolutions; things like exercising when you don't enjoy it, eating uninteresting but healthy food, or giving up smoking all have potential good outcomes in the distant future, but bring little enjoyment in the present.

The same goes for speeding: The thrill of going fast is in the present, and the reduction in time spent on an unexciting journey, or perhaps the removal of the possibility of arriving late at one's destination, is in the immediate future.

Losing money as a result of a traffic fine is a much more distant, much less certain outcome. Think of all the times you've sped or run a red light without getting a fine.

There are two ways to reduce the likelihood of any behaviour occurring.

We can punish bad behaviour when it happens – by issuing fines, confiscating cars and licences, or sending people to jail.

We can also reward alternative, desirable behaviours.

These strategies for behaviour change aren't mutually exclusive; we can punish bad behaviour and reward good behaviour in tandem.

Nevertheless, society has a preference for punishment alone, perhaps because it is easy, satisfying, and (at least in the case of fines) apparently lucrative.

Yet it's hard to get punishment right; research has shown that even immediate, aversive consequences such as a strong electric shock don't always stop the behaviour that produces them.

Under particular conditions, electric shocks can actually increase the behaviour they are intended to punish.

Another problem with punishment is that we often feel resentful about doing things to avoid a bad consequence.

This means we're not likely to do those things when we think we won't get caught, and that we're likely to form negative impressions with anything that's linked with those bad consequences occurring (like the police who issue fines).

In contrast, we are highly motivated by potential reward – money, recognition, social approval.

Think of the things you do because you want to (not because you feel you should) – they're usually things that lead to something good.

We're so motivated by potential reward that we'll often try to obtain a reward even when it's not highly likely to happen.

Reward is a powerful way to change behaviour, but it's not often used on a societal level.

Think about your experience as a driver. When was the last time you were rewarded for driving at the speed limit?

With the 2017 road toll at an unsettling high, it's clear we need to do more to encourage people to drive safely.

Dr Gilbert is right: We do need to have a conversation about fines.

But this conversation shouldn't be about how fines are more or less serious for people in particular income groups; it should be about why people keep getting fines – why fines aren't having a meaningful impact on dangerous driving.

Ineffective interventions are a huge cost to society; think of the manpower involved in catching speeders, issuing and recording fines, making sure fines get paid, and dragging people through the justice system when fines don't get paid.

Where ineffective interventions target behaviours that cause serious harm (and death) – like potentially dangerous driving – the costs are more serious than mere financial burden. Just ask the families of victims of traffic accidents.

The income-related unfairness Dr Gilbert discusses is a symptom of a much more serious problem: Our existing methods of making people drive safely don't seem to work particularly well. People keep getting fines when they can't afford them because fines don't stop bad behaviour.

Dr Gilbert's conversation needs a new direction: What can be done to make people more likely to drive safely on our roads?

Research tells us that delayed, low-probability punishments (like fines) have very little impact on our decisions. What alternatives to punishment do we have?

What would happen were we to use revenue from fines to fund cash rewards for drivers who stick to the speed limit, to offer a discounted car registration to drivers who drive safely, or to reduce ACC levies for employers whose employees drive safely?

Like punishment, reward can be cost-neutral.

As ridiculous as these suggestions might sound, at the very least, they might be the next verse in Dr Gilbert's conversation about how we address the serious limitations of fines.