At a presentation on Wednesday 6th May, Kela, the Finnish social security agency, gave further results from the first year of its Basic Income experiment.
The trial group was 2,000 randomly selected unemployed individuals who had their unemployment benefit made unconditional for a period of two years. A control group of 173,000 unemployed individuals had no changes made to their unemployment benefit.
During the first year of the trial there was no statistically significant change in employment market activity among the trial sample. Analysis of the data generated by the second year of the experiment has now shown that, for the trial group, employment rose on average by six days between November 2017 and October 2018. Larger increases were experienced by families with children, and by individuals whose mother tongue was not Finnish. Evaluation of the second year’s employment data had been complicated by the implementation of a more activation-oriented social security system for unemployed individuals half way through the experiment, which means that changes in employment market behaviour will have been affected by various consequences of the new policy as well as by the unconditionality of the trial group’s unemployment benefit.
The response rate to survey questions about wellbeing was predictably low, but it had still been possible to conclude that, compared with 5,000 randomly selected individuals from the control group, the trial group had experienced a higher rate of generalised trust, less stress, less depression, less bureaucracy, less financial stress, and better cognitive functioning.
From interviews with 81 recipients of the Basic Income, it was discovered that some had experienced a wider variety of participation in society outside employment, and that a sense of autonomy had increased.
The researchers had concluded that wellbeing effects were more significant than employment market effects, which mirrored results from experiments with different but similar mechanisms in Canada and the Netherlands.
A telephone survey to gauge public opinion after the experiment had found that 46% of respondents believed that a Basic Income should be introduced.
The discussion that followed the presentation explored the definition of Basic Income, whether different experiments could be compared if they were experimenting with different things, the importance of a secure layer of income, how long it would take to implement a Basic Income, the importance of social experiments, whether a Basic Income would make people lazy, and the extent to which the effects of a nationwide and permanent implementation of a Basic Income scheme would differ from those of a two year experiment.
Still to do: a report in English; a report on experiment participants’ use of other social benefits and services; and a study of the reasons for individuals with a non-Finnish mother tongue had been disproportionately enabled by their Basic Incomes to gain new skills and find employment.
To see a recording of the presentation, click here.
The final report can be found here. An English summary will be found on the last few pages, starting on page 187.
Social Europe has published an article about the results by Philippe Van Parijs
.
Now it is clear that the Finnish basic income experiment has shown both well-being effect (which was already known) and employment effect (which was uncertain until now).
The difference between the treatment and control groups in the number of days in employment during the first year was only 0.39 days, but there was a 6-day difference during the second year (November 2017-October 2018, exactly).
Kela has indicated, “the interpretation of the effects of the experiment is made more complicated by the introduction of the activation model at the beginning of 2018, which meant more stringent entitlement criteria for unemployment benefits asymmetrically in both groups.”
This can be interpreted that the basic income was more effective than the activation model (more stringent monitoring and sanctions) in encouraging the unemployed to find a job.
In fact, monthly reports during the first year suggested that employment effect was nil during the first nine months but seemed to occur starting in October 2017. It seems that it took some time for the employment effect to materialize. It is hard to imagine that the employment effect was mainly caused by the activation model that was introduced in January 2018.