Under ‘Mutual Obligations’, the Government requires the unemployed to complete activities such as attending appointments with their employment services provider, searching and applying for jobs and participation in work-for-the-dole.
In return, the unemployed receive job search assistance and unemployment benefits. The latter is conditional, since failure to comply with ‘mutual obligations’ requirements may lead to (partial) suspension of unemployment benefits.
Although such a penalty may lead to more compliance with the requirements, as suggested by Andrew Wright in a recent TTPI blog entry, the proof of the mutual obligations pudding is not in the compliance.
The proof is in its ability to assist the unemployed to return to work as quickly as possible.
Materials and methods
We make use of the Household, Income and Labour Dynamics in Australia (HILDA) survey, Australia’s most comprehensive longitudinal dataset in this field of research. Data collection started in 2001. We now have 19 (annual) waves at our disposal.
We focused on unemployed respondents in the survey and split the group in two parts: those who indicated in a particular wave that they had to fulfil ‘mutual obligations’ requirements and those who did not.
In terms of labour market outcomes, we focused on five indicators.
We determined the number of hours the unemployed person spent on search for work. Then we used information from the subsequent wave, to determine their employment status 12 months later and the qualities of the job they held at that time if they were employed. We focused on hourly wages and hours worked.
Since respondents fill out a monthly employment calendar in each annual wave for the preceding twelve months, we were able to determine how many months it took them to find employment (time-to-employment) and how many out of the 12 months they spent in employment (time-in-employment).
A straight comparison of performance on these labour market outcomes of the two groups is problematic. Take, for example, unemployment duration (time since the last job).
Unemployment duration raises the likelihood of having to meet a ‘mutual obligations’ requirement but is also known to reduce the chance of reemployment. So, the group that faces the ‘mutual obligations’ requirement is likely to experience longer unemployment duration than the group without. And unemployment duration matters for labour market outcomes.
Therefore, to isolate the effect of the ‘mutual obligations’ requirement, we need to control for unemployment duration differences between the two groups. The same applies to any other factors that may jointly determine the likelihood of having to meet a ‘mutual obligations’ requirement and labour market outcomes.
In total, we include 42 variables in our analysis, which we believe may be such joint determinants. These variables include (for a full list, refer to the supplemental material of the full research publication):
- Unemployment history (for example, time in unemployment as share of time since completing full-time education, time currently in unemployment, unemployment benefit recipient)
- Time-variant heterogeneity (for example, (changes in) household income, debt, composition, homeownership)
- Time-invariant heterogeneity (for example, personality traits, gender, native speaker of English)
- Labour market environment (for example, regional unemployment rate, net replacement rate, underutilisation rate, trade union density)
We then conduct a matching analysis.
The matching analysis compares labour market outcomes for those with a ‘mutual obligations’ requirement to those looking for work without such a requirement, but who are otherwise very similar (similar profile with respect to the 42 control variables).
Joint determinants, which obfuscate the analysis if not controlled, now no longer play a role.
However, there is another concern. ‘Mutual obligations’ requirements are not handed out randomly.
There are set criteria. One can add these criteria to the list of control variables in the matching analysis, but if we do so, we would struggle to match those with and without a ‘mutual obligations’ requirement. That is, these two groups necessarily differ in terms of meeting the criteria for the application of a ‘mutual obligations’ requirement.
Except, if the criteria have changed over time, or the criteria have not been applied correctly all the time. Both occurred and we exploit these exceptions. That is, either side of a rule change one can find unemployed persons who meet the latest criteria for a ‘mutual obligations’ requirement, but they only apply to one group; not the other.
Findings
Our research shows that a ‘mutual obligations’ requirement does not affect hours spent looking for work.
Instead, those with a ‘mutual obligations’ requirement needed more time to find a job (time-to-employment) and spent less time in employment in the 12 months (time-in-employment) that we followed them.
After 12 months, the ‘mutual obligations’ requirements had no effect on the likelihood of being employed but adversely affected the quality of that employment. Those who looked for work with a ‘mutual obligations’ requirement held jobs with lower hourly wages and fewer hours, which translates to lower weekly wages.
We find that ‘mutual obligations’ do not pass the test of sound labour market policy. Hence, we recommend abolishment.
‘Mutual obligations’ requirements seem to hinder rather than help the unemployed return to employment. This is not, because time spent towards fulfilling ‘mutual obligations’ requirements is time not available for job search. We find no differences in the number of job search hours between the two groups.
Quality of job search
The reasons for our findings are likely related to the quality rather than the quantity of job search.
Since job search quality is hard to quantify, economists rarely investigate it. We therefore broaden our horizon, in line with Simone Casey and David O’Halloran’s suggestion in a recent TTPI blog entry.
In our article, we hypothesise three potential explanations for our findings stemming from the cognitive domain, which argue that the application of and financial threat arising from a ‘mutual obligations’ requirement:
- changes the motivation of the subjected unemployed person to search for work from intrinsic to external (self-determination theory);
- causes stress, which prevents the unemployed person from accessing the part of the brain which is responsible for careful, deliberate decision-making (scarcity theory);
- forces the unemployed person to spend cognitive resources on meeting the ‘mutual obligations’ requirement and associated threats, which are then not available for job search (conservation of resources theory).
All three theories predict that a ‘mutual obligations’ requirement erodes the effectiveness or quality of job search and are therefore potential explanations for our findings.
A macroeconomic problem
Abolishment of ‘mutual obligations’ does not solve unemployment though. The causes of unemployment are macroeconomic in nature.
The Australian Bureau of Statistics (ABS) estimates that 701,100 people are unemployed (seasonally adjusted) in May 2021. At the same time, the ABS estimates 362,500 vacancies (seasonally adjusted). That means for every vacancy, Australia has 1.9 unemployed persons (leaving the underemployed out of the analysis, including them would make matters even worse).
Evidently, there is not enough work for all that want work.
Labour market policy instruments that essentially treat unemployment as a microeconomic problem (This means unemployment is a problem of the unemployed persons themselves), ignore this macroeconomic reality.
Microeconomic studies often show how changes to policy settings (sticks and carrots) can improve the job find success of the unemployed they study. But in an environment where there are not enough jobs to go round, that individual success is likely to come at the expense of someone else, who remains outside the scope of the study.
We therefore advocate governments first and foremost study unemployment as a macroeconomic phenomenon and create enough employment to satisfy the employment needs of all. The best way to do that is to introduce a job guarantee.
The full research publication is available Open Access:
Ruud Gerards and Riccardo Welters (2021). Does eliminating benefit eligibility requirements improve unemployed job search and labour market outcomes? Applied Economics Letters, DOI: 10.1080/13504851.2021.1927960
Thank you both for an excellent research study. Your findings reinforce the long and painstaking research work of Bill Mitchell et all others at the Centre For Full Employment and Equity at the University of Newcastle and the work of his American Modern Monetary Theory colleagues at The University of Missouri Kansas City and The Levy Institute