Photo by João Ferrão on Unsplash https://bit.ly/39tdsrv

Being unemployed is difficult, and unemployment payments are designed to ensure that job seekers are generally financially better off from increasing their hours of work. So one might assume that benefit recipients would do everything they could to find work as quickly as possible – and any requirements to do so would have little effect on this tendency.

However, there is a large volume of evidence that compulsory requirements, such as job search, speed re-entry into employment – including during recessions. Moreover, by definition, compulsory requirements involve some consequence for not meeting those requirements. Typically internationally, this includes financial penalties in the form of reduction or cancellation of unemployment payments for a period.

A growing body of research has shown that the application of penalties has a positive impact, and can as much as double the rate of benefit recipients finding work. However, there are currently gaps in the academic evidence.

For example, the assumption that higher penalty values lead to more behavioural change has previously been difficult to test. This is because although many countries have different penalty values, they are typically for different types of failure or apply to different groups.

It is therefore challenging to determine that the difference in penalty value was what changed behaviour, and not some other factor.

Do higher penalties change behaviour?

Despite this, many countries’ unemployment payments have severely escalating penalty values to deter unwanted behaviour.

For example, the United Kingdom’s Universal Credit payment has four categories of penalty type (lowest, low, middle and higher), depending on the severity of the requirement missed. Furthermore, there are escalating penalties of up to 26 weeks without payment for a second instance of non-compliance.

Implicit in this is an assumption that the main driver for benefit recipients to change their behaviour is the direct penalty value. However, increasing penalty values is likely to have diminishing returns, and is unlikely to be effective past a job seeker’s ability to afford the penalty.

To use an analogy perhaps more familiar to some people, a driver with only one demerit left on their license is unlikely to change their behaviour further when double-demerits are announced.

There is also evidence internationally that severe penalties for the unemployed can lead to serious negative unintended consequences such as increased exit from the labour force, increased crime, increased food bank use, and unintended costs on other public agencies, charities and voluntary organisations.

While the above evidence is from jurisdictions with generally much larger penalties for not complying with requirements than Australia, it highlights that there can be both positive and negative effects from penalties.

Evidence is therefore needed to be able to determine appropriate penalty values which balance the objective of encouraging behavioural change while minimising the above unintended consequences.

My co-authors and I recently examined this issue – first by explicitly looking at the effect of changing penalty values, and then by looking at some of the administrative elements associated with a penalty that generally do not have a financial impact.

In Australia, previously job seekers’ penalty value changed based purely on appointment availability

Under Australian penalty arrangements that existed before July 2018, if a job seeker did not attend a compulsory appointment with their employment service provider (who administer job seekers’ requirements) without a valid reason, the provider could recommend a penalty apply.

Penalties were calculated as 10 per cent of the basic income support payment rate multiplied by the number of days until the job seeker attended a rescheduled appointment.

If an appointment was not available within two business days, then no penalty would apply. However, where an appointment was available within two business days, job seekers could receive a 10 per cent or 20 per cent penalty based purely on appointment availability, which is outside of their control (usually in the order of magnitude of $50 or $100).

To determine the extent to which job seekers’ behaviour was affected by this change in penalty value, we examined job seekers’ likelihood of attending their next appointment after matching their modelled likelihood of receiving a one or two-day penalty using demographic information contained in 2015-16 administrative data and a propensity score matching approach.

While it is unusual to look at the effects of unemployment payment penalties on future compliance (most studies examining unemployment payment penalties look at employment outcomes), we do this because job seekers are penalised for not meeting requirements rather than not finding work. So we consider that it makes sense to study the extent to which penalties affect compliance.

Higher penalties increased attendance at the next appointment

We found that a two-day penalty relative to a one day on average increased attendance at the next appointment of 2.4 percentage points.

This may not initially seem like a large increase in attendance, but these particular penalties were quite rarely applied in practice. Typically they were only applied to serially non-compliant job seekers (who on average only attended 61.2 per cent of appointments). So in this context, a 2.4 percentage point increase is not insignificant.

However, it is also not only the financial impact of penalties that could have an impact.

For example, penalties may increase job seekers’ awareness of what is required of them or their expectation of the likelihood of a penalty being applied. Similarly, job seekers who have been penalised may be monitored more strictly (or believe themselves to be), or feel an element of stigma. All of these factors may change job seekers’ behaviour.

Moreover, while ordinarily, it is challenging to study the effects of these elements of penalty application without results being affected by the financial impact of penalties, Australia has a unique system of temporarily pausing job seekers’ payments which allows this.

Splitting the administrative and financial effects

When a person does not meet a requirement, their payment may be paused (suspended) until they re-engage with their requirements, at which point payment is released with back-payment of any delay.

While a delay in payment may hurt job seekers (who are unlikely to have access to significant cash reserves or credit), most payment suspensions do not result in a delay. This is because benefits are paid fortnightly, and suspensions can happen anytime during that fortnight. In other words, if the duration of a suspension does not coincide with a person’s payment date, then they experience no delay in payment. Moreover, because suspensions are typically short, most (86 per cent) result in no delay in payment.

This means that for most suspensions, while job seekers experience the administrative elements of a penalty, they do not experience any financial impact. In addition, before the new arrangements in 2018, when a person did not meet a requirement, providers could decide to use ‘discretion’ and effectively ignore the failure.

To test the impact of these generally zero-dollar financial penalties, we used 2015-16 administrative data, and the propensity score matching approach matched job seekers’ likelihood of facing a suspension or their provider using ‘discretion’ to ignore the failure.

We found that where the job seeker had experienced a suspension, their likelihood of attending subsequent appointments increased by 13 percentage points. Interestingly too, they were more likely to attend subsequent appointments as well.

Consistent with previous research internationally on penalties, we found that women changed their behaviour more in response to suspensions than men. We also found that job seekers with the lowest number of previous instances of non-compliance were most likely (22 percentage points) to change their behaviour following suspension.

However, interestingly even our most non-compliant group (those with 27 or more previous events of non-compliance) were 9 percentage points more likely to attend their next appointment following a suspension.

Temporary payment suspensions do have effect, but…

Our results suggest temporary payment suspensions, like those in the Australian payment system, are effective at encouraging job seekers to comply with their requirements. This may provide a policy avenue for other countries to reduce the potential adverse effects of strong penalty application.

However, the fact that that some job seekers in our 2015-16 data were significantly non-compliant arguably demonstrates that suspensions alone may not be enough and that larger penalties may be appropriate for serially non-compliant job seekers.

To this end, our research has also shown that higher value penalties do change behaviour more – at least for the cohort we studied, and at the values in our study.

Our research, as well as earlier international work, has shown that some cohorts are more responsive to penalties than others. However, this is a policy area which would benefit from more extensive study to allow policymakers to achieve desired behavioural effects while minimising potentially adverse severe unintended consequences.

Comments are closed.