The Finnish government conducted a basic income experiment from 2017-2018 to study its potential effects. The experiment provided €560 per month to a random sample of 2,000 unemployed individuals, while a control group of 170,000 unemployed individuals received standard benefits. Preliminary results after the first year found no significant impacts on employment but significant improvements in wellbeing for the basic income group. More conclusive results will be available after further analysis and upon completion of the two-year experiment in 2020.

1. Finnish Basic Income
Experiment 2017-2018 -
About the experiment and
its’ evaluation
Minna Ylikännö, Ph.D.
Senior Researcher
Kela – the Social Insurance
Institution of Finland
minna.ylikanno@kela.fi

2. Background
The Center-True Finns-Conservatives coalition cabinet (nominated 28. May 2015)
took basic income (BI) experiment in its working program by referring to:
• Changes in the labor markets
• Does the Finnish social security system properly correspond to changes in the labour market?
– High level of structural unemployment, automatization, robotization etc.
• Elimination of incentive traps in the social security system
• Too many cases where work does not pay (enough)
• Elimination of bureaucratic traps
• Need for a more transparent and less complicated system instead of “social security jungle”
• Main interest of the Government was in the possible positive effects of the BI on
the employment rate

3. Steps towards the experiment…
• €20 Mill. for the experiment
• Some extra funds for the planning of the BI experiment
• Open competition on the funds
• 15. September 2016 Kela’s consortium was selected to plan the
experimental setting and the model(s)
• Work began in the mid-October 2015
• The first report was delivered 30th of March 2016
• The final report delivered the 16th of December 2016
• The experiment started 1.1.2017 and lasts for 2 years

4. In the planning process it was to study
• Which kind of models are most suitable for the experiment?
• What should be the level of the monthly payment
• How to combine BI with income-related benefits and other basic
benefits?
• How should the taxation be taken into account in the different
models?
• What are the strengths and weaknesses of different models in the
context of the EU legislation and the Finnish Constitution?
• Also, it was required that the researchers give recommendations
on the experiment – what should be the model(s) to be
experimented?
4

5. Models explored and developed
• Full basic income (BI)
• The level of BI is high enough to replace almost all earnings-related benefits
• The level of monthly payment has to be high, €1 000-1 500
• Partial basic income
• Replaces all ’basic’ benefits but almost other benefits left intact
• Minimum level should not be lower than the present day minimum level of
basic benefits (€550 – 600/month)
• Negative income tax
• Income transfers via taxation system
• Other models
• E.g. low level of BI added with some kind of ’participation’ income
5

6. The experimental setting planned by the expert group
• The entire adult population (excl. pensioners) is used as a basis for the
sample
• age and income selection criteria
• low-income earners
• Between 25 and 63 years of age
• Weighted sample of particularly interesting groups
• Nation level randomization to get representative results
• Local experiments in order to capture networking, institutional and
interaction effects and externalities
• In order to have a sufficiently high sample size, Kela benefits would be
used as a source of extra funding (sample size could be as high as
10,000 persons)
•
6

7. What was experimented?
• BI €560€ a month (tax-free)
• Present taxation on income
exceeding 560€
• Social benefits exceeding 560€ were
paid out as previously
• Housing allowance and social assistance
were tested against basic income
• Work income on top of BI without
tax “penalties”
• 2 000 unemployed who received
flat-rate unemployment benefit
from Kela in November 2016
• Random nation-wide selection into
the treatment group
• The rest of the unemployed
receiving benefits from Kela (app.
170 000) form the control group
• Obligatory participation
• BI experiment began 1.1.2017
and ended 31.12.2018
7

8. Why the experiment ”shrank”?
• Constitutional constraints
• Question on equal treatment
• Tax authorities had not enough time to
change tax laws for the experiment
• Tax-free BI combined with present tax
system
• Only unemployed who received
unemployment benefits from Kela were
selected to the experiment
• Easy to make a random nation-wide
sampling
• Easier to write legislation for one specific
group than for many heterogeneous
groups
• Other legal constraints
• Implementing BI in a complex institutional
setting is very demanding
• Time pressure
• There was little time to write and pass the
legislation
• Also, there was not enough time to create a
separate ICT platform for paying out the
benefit
• Creating proper ICT systems for payments
limited the size of the treatment group
• Partially manual decisions and payments
8

9. How the experiment is to be evaluated?
• Before the experiment it was decided that the receivers of BI are not to be contacted by
the researchers
• No surveys or interviews during experiment
• The main interest is in the changes of employment and income
• Registers are the main source of information – combined registers from various administrative
organisations
• Secondary outcomes will be studied via surveys and interviews
• Economic stress, general well-being, health, social relations, experiences on bureaucracy etc.
• First results was published in 8.2.2019 – the report includes register based analysis for
the year 2017 and preliminary results from a phone survey collected in the end of year
2018
• The second report on the results will be published 4/2019
• The final results will be published in the early 2020
9

10. The preliminary results of the BI experiment
• Multidisciplinary research project utilizing both quantitative and qualitative data
• All the results are based on a comparison of the experiment group and the control
group and on statistical testing of the differences
• The results are preliminary insofar as the register data at this stage only cover the
first year of the experiment, 2017
• We are thus unable to analyse the effects of the experiment on employment status and
other behaviour for the whole experiment
• The survey covers both years of the experiment, 2017 and 2018
• Later, by combining register data with the survey - in this way we obtain more reliable
results as regards the wellbeing effects of the BI experiment
• We will also utilize the information gathered in the interview study conducted in the
research project
14.2.2019 Yksikkö/Projekti/Esittäjä | Esityksen otsikko10

11. Employment effects of the BI experiment
• The recipients of a basic income were no better or worse at finding employment than those
in the control group during the first year of the experiment (2017) - there are no statistically
significant differences between the groups
• The recipients of a basic income had half a day more of employment in the open labour
market than the control group
• Having earnings from the open or the subsidised labour market was more frequent among
the recipients of a basic income than in the control group by one percentage point
• However, earnings and income from self-employment were on average 21 euros lower in
the experiment group than in the control group.
•  BI experiment did not increase employment in the first year of the experiment
14.2.2019 Yksikkö/Projekti/Esittäjä | Esityksen otsikko11

12. Wellbeing effects of the BI experiment
• Wellbeing of the recipients of a basic income was clearly better than
that of the control group
• Those in the experiment group experienced fewer problems related to health, stress,
state of mind and ability to concentrate than those in the control group
• Those in the experiment group were also considerably more confident in their own
future and their ability to influence societal issues than the control group
• Generalised trust, i.e. trust in other people, was on a higher level in the experiment
group
• There was almost no difference between the groups as regards trust in different
institutions, such as the court system and the police
• However, recipients of a basic income trusted politicians more than did the control
group.
14.2.2019 Yksikkö/Projekti/Esittäjä | Esityksen otsikko12

13. Contradictory results?
• On the basis of the register analysis, there is no difference between the groups as
regards employment status - the survey results however showed significant
differences between the groups for different aspects of wellbeing
• Also, those in the experiment group were more confident of their employment
prospects than the control group and replied that the basic income reduced the
bureaucracy involved when claiming social security benefits and when looking for a
job and that the basic income made it easier to start a business
• One could argue that the results are contradictory – this is however not the case:
even if the basic income had no effects on employment status one way or the other,
it may still have significant effects on wellbeing
• This is in line with the findings of the experiments implemented in other countries
(e.g. Canada) – wellbeing effects are more pronounce than the employment effects
14.2.2019 Yksikkö/Projekti/Esittäjä | Esityksen otsikko13

14. More results later this year – final results in 2020
• The results presented in this report are preliminary - one should not draw any
firm conclusions about the effects of the basic income experiment on
employment status and wellbeing
• We can report on the real effects of the experiment in a reliable way only
when all the materials for the evaluation study - register data, surveys,
interviews and different combinations of these - have been analysed in more
detail and the effects of the political, institutional and schedule-related
parameters that created a framework for the experiment have been evaluated
• Only when we have the final results of our research projects, conclusions can
be made concerning the implications that an introduction of a basic income
could have on individual labour market behaviour and wellbeing in Finnish
society.
14.2.2019 Yksikkö/Projekti/Esittäjä | Esityksen otsikko14

15. Thank you!