Universal basic income is not unemployment’s silver bullet

Discussions around universal basic income have sparked global enthusiasm for the notion. The fight for social justice dominates its proponents’ discourse on the matter, while policymakers remain more interested in its alleged positive effects on employment and income generation. With automation and robots poised to destroy jobs but create work, would basic income help to secure employment in the future?

The concept of basic income is not a new, or untested, one. It first attracted interest from American and Canadian legislators and governments in the 1970s, leading to local experiments. However, Finland was the first country to implement a nation-wide randomised field experiment on the idea.

From the 1970s onward, interest around basic income faded. Social policies on both sides of the Atlantic have since prioritised ‘workfare’, under which payment of benefits is made conditional on participation on job-promoting activities such as training, rehabilitation and work experience, or on unpaid or low-paid work. But while the number of people relying on these policies has grown, the number of available jobs has not.

The Finnish experiment concerned itself with partial basic income and targeted able-bodied people without work. Thus, it was not an evaluation of universal basic income. A group of 2,000 individuals were randomly selected from those who, at the end of 2016, had received a flat rate unemployment compensation. These were young unemployed people without work history and long-term unemployed people who had exhausted their right for earnings-related unemployment compensation. The experiment group was paid a net sum of 560 euros per month for a study period of two years ranging from 2017 to 2018, regardless of employment status. The rest of the unemployed with flat rate unemployment compensation served as the control group. Participation was mandatory under a law which was passed to implement the experiment.

Since the experiment group getting the ‘treatment’, i.e. basic income, was similar to the control group in all relevant background characteristics, the experiment mimics experiments carried out in the natural sciences and in medicine. If there were differences after the experiment between the treatment group and the control group, we may establish a causal loop.

In February, the basic income research team released the first results of the experiment. The first year’s results showed that the basic income recipients did not have more work days or higher incomes than those in the control group. Despite the fact that basic income recipients had better incentives to work, there were no statistically significant differences between the groups.

The research team also released results from a survey conducted at the end of last year. The survey showed that basic income recipients experienced significantly fewer problems relating to health and stress. They also had a more positive view of their future.

The partly conflicting results led to two competing narratives. The first camp pointed to disappointing employment outcomes, while the second camp emphasised positive subjective perceptions among the basic income recipients. The latter camp also highlighted that the experiment group did not work less during the first year of the experiment.

The experiment mostly set out to study how the social security system could be reshaped in a way that promotes active participation and gives people a stronger incentive to work. That means employment outcomes were the focal point of the endeavour. The idea was not to study if basic income would decrease labour force participation. For that purpose, employed persons, particularly workers in low paid and precarious jobs, would have been included in the experiment.

The same goes for subjective well-being. It was not a topic of original interest in the experiment. There was no baseline survey conducted on subjective well-being to act as a point of analysis. In other words, the survey results do not stem from the same randomised control trial setting as the employment results. While the employment outcomes are measuring the effect of basic income on employment, that is to say the change caused by basic income, we do not know if people’s subjective assessments changed after they started to receive basic income.

The positive evaluations may not relate to basic income but they do have bearing on the public debate that surrounds basic income and to the fact that people were members of a selected group. The survey results are also undermined by a low response rate: only 31% in the experiment group and 20% in the control group partook in it. The survey was carried out in late 2018 when the newly implemented and heavily criticised sanction scheme was implemented for the control group. This alteration may have also had implications for the results, especially if only basic income recipients with positive attitudes participated in the survey.

Finally, it is important to note that behaviour and attitude are not equal outcome measures. A number of biases may distort surveyed attitudes while behaviour, in this case employment, reveals actual preferences.

In the debate surrounding the Finnish experiment, perceived well-being is understandably emphasised by those who see basic income more as a social justice issue than as an instrument for activating unemployed people. However, even those that find themselves in this camp should acknowledge the fact that the negative employment effect observed during the Finnish experiment is more robust than the positive effect it had on subjective well-being.

For those who argue that basic income will easily spur people at the lower ranks of society to seek self-reliance, the preliminary results from Finnish experiment come as a slap in the face. The main message from the experiment is that the long-term and young unemployed face other obstacles to work than financial incentives. Barriers such as social problems, health issues, low education and outdated skills prove to be more limiting factors. This finding shows that we need tailor employment promoting services as we encounter future labour markets.