Ville-Veikko Pulkka is a Doctoral Researcher at the University of Helsinki. He has previously worked with the Finnish Social Insurance Institution Kela on their recommendations for the Finnish basic income experiment.
The Finnish basic income experiment, the first nationwide randomised controlled trial testing the effects of an unconditional basic income, reached its halfway point on 1 January 2018. Given the unique position of Finland in the present basic income discussion, the country makes a particularly interesting case for studying the public view on basic income. Based on nationally representative survey data from autumn 2017, this blog post investigates whether basic income has real prospects to become a politically feasible option after the experiment is complete.
Methodological pitfalls of basic income surveys
Since 2015 four different research organisations have conducted surveys measuring support for basic income in Finland. The measured support has varied from 29% to 79%, which reflects the many methodological pitfalls involved in basic income surveys.
The highest support rate (79%) for basic income thus far was measured by the Centre party’s think tank e2 in 2015. The high support rate is unsurprising when the definition of basic income in this survey is considered: in this study basic income was defined as a benefit that automatically incentivises working and entrepreneurship.
If the survey by e2 speculates as to the favourable consequences of implementing a basic income, another survey by Finnish Business and Policy Forum EVA from 2017 takes the opposite approach. The survey carried out by EVA framed basic income in terms of its potential problems, suggesting that implementing the policy at the level of minimum income (understood in Finland as the level of social assistance, €491.21 a month) would require tightening income taxation[i]. This survey found support for basic income was only 39%.
Between these two studies, in 2015, the research section of the Finnish Social Insurance Institution (Kela) conducted its own basic income survey. The survey by Kela measured support of 69% for basic income. The study did not speculate as to the conceivable effects of basic income, but only stated that basic income is a benefit that is “guaranteed automatically to everyone”. Nevertheless, one can ask whether this definition is unambiguous enough. I posit the view that it may not be, since the fourth survey – published by the European Social Survey in late 2017 – measured support of 59%, and the definition of basic income was more accurate, emphasising the unconditional nature of basic income.
Apparently, the methodological pitfalls of basic income surveys had been considered by my former colleagues at Kela, as the second section of the above-mentioned Kela survey attempted to tackle biases resulting from overly general definitions of basic income. When various levels of basic income were combined with speculative flat-rate taxes, the support collapsed from 69% to between 29 and 35%. Still, even this approach raises questions on the generalisability of the results. I strongly believe most people don’t realise that the relatively high marginal flat-rate tax presented does not change their net taxation (ie. when basic income is considered) significantly. Therefore, including taxation in basic income surveys may also lead to biases. Moreover, flat-rate taxation is generally a non-starter in the context of a Nordic welfare state that has traditionally relied on progressive taxation.
A more realistic view on basic income’s support in Finland
To tackle the methodological pitfalls of the previous surveys, yet another survey was necessary to estimate the support for basic income in a reliable manner. A survey[ii] conducted by my colleague Professor Heikki Hiilamo and I in autumn 2017 focused mainly on Finns’ views on the future of work and preferred policy responses to the digital economy, but also included a separate section on basic income. Due to the limited number of questions, basic income could not be studied as comprehensively as one might have wished. Nevertheless, I argue that our survey succeeded in providing a realistic view on basic income’s support in Finland.
To provide reliable measurements[iii], the characteristics of basic income were described to respondents based on the generally accepted definition from the Basic Income Earth Network (BIEN). Additionally, the level of basic income and the most significant non-replaceable benefits were explicitly defined. The net level of €560 a month became an essential parameter since this level approximately corresponds to the net level of many basic security benefits in Finland. This is also the model that is now tested in the Finnish basic income experiment.
Table 1 provides an overview of support for six various basic income models and participation income[iv]. Participation income is not a basic income model by definition, but since it has often been proposed as a more feasible alternative to basic income, it is included in the overview.
Table 1. Support for six various basic income models and participation income in Finland
|Basic income model||Good idea||Neither good nor bad idea||Bad idea|
|Partiala basic income of €560 a month||51%||20%||21%|
|Partiala basic income > €560 a month||33%||20%||39%|
|Partiala basic income < €560 a month||27%||27%||37%|
|Fullb basic income of €1500 a month||25%||17%||66%|
|Partiala basic income of €1000 a month||24%||17%||51%|
|Fullb basic income of €1000 a month||20%||20%||51%|
a Maintains eligibility for housing allowance and earnings-related benefits.
b Withdraws eligibility for housing allowance and earnings-related benefits.
c Eligibility for social assistance and basic security benefits requires participation in activation measures that can be defined by the unemployed in a more autonomous manner than currently (eg. voluntary work, studying, caring for close relatives or leisure activities).
As Table 1 shows, Finns find the model that is currently tested, a partial basic income of €560 a month, the most feasible option among the six presented models. Nevertheless, in comparison with the previous surveys, and given the broad endorsement for participation income, the support for basic income is rather moderate at best. Neither weakening nor strengthening the level of social security with a basic income gathers considerable support. Since a thorough investigation of differences in sociodemographic characteristics is out of the scope of this blog post, I shall focus only on the most popular model.
Essential differences in sociodemographic characteristics
Interestingly, age was observed to have a clear connection to the support of basic income. As can be seen in Table 2, the age group from 15 to 24 reported significantly more support for basic income in relation to the other age groups. This result is significant at the p = 0.000 level.
Table 2. Support for a partial basic income of €560 a month by age group
By occupational group, the most dedicated support for the analysed model was provided by lower clerical workers and pupils/students. The results obtained from the analysis by occupational group are presented in Table 3 (p = .000).
Table 3. Support for a partial basic income of €560 a month by occupational group
|Employee||Lower clerical worker||Upper clerical worker||Manager||Entre-preneur||Pupil/
Perhaps less surprisingly, basic income is relatively popular among the unemployed. The differences by labour market status are highlighted in Table 4 (p = .020).
Table 4. Support for a partial basic income of €560 a month by labour market status
|Labour market status||Total|
|Full-time employee||Part-time employee||Unemployed||Outside workforce|
The first nationwide basic income experiment launched by Prime Minister Juha Sipilä’s centre-right government in January 2017 has attracted broad international interest since the first plans became public in 2015. What has repeatedly been dismissed in the media hype is the context of the experiment: basic income is tested in addition to several other social experiments, and the politically determined focus is mainly on labour supply. This implies that the experiment is not a break with the activation policies implemented since the mid-1990s in Finland. This rather evident observation for a Finnish scholar was manifested in January 2018 when the government implemented a controversial active model for the unemployed. In brief, the active model translates into more reporting obligations, conditionality and sanctions for the unemployed. Regardless of broad criticism directed at the active model, the government has already made further plans to tighten up the conditions of unemployment benefits.
To put it differently, basic income has a huge burden of proof to bear if it is to become a politically feasible option – even in the context of Finland. Within the activation paradigm this means that to convince economists and politicians, basic income experiments ought to provide strong evidence that basic income increases labour supply considerably. Since the government has not committed to budget the experiment in a manner that would enable testing a basic income with a representative population, the results can only be generalised from those we see with the long-term unemployed and people with a short working history. Our final report for the experiment from December 2016 emphasised the need for an extension of the experiment, but given the described context, it was unsurprising that the government did not follow the recommendations of our research group.
It is evident that Juha Sipilä’s government is committed to pursuing activation policies relying on increasingly conditional benefits. However, the findings of our survey suggest that the Finnish citizens are also somewhat convinced that there is no need for a paradigm shift. This is apparent not only from the moderate support for basic income, but also from the fact that according to our data, Finns tend to support conventional policies emphasising labour supply and employability. These include policies such as increasing economic incentives to participate in the labour market, investing in education, increasing the current activation measures and moderately, but just moderately, reducing means-testing.
At present, the most dedicated supporters of basic income are the youth and the unemployed. This indicates that precarious positions in the labour market have an explicit connection to willingness to endorse basic income. Similarly, lower clerical workers, according to our data the group most concerned about the future of work, are supporting basic income more often than the other occupational groups. Therefore, our data suggest that should unemployment, underemployment or precarious jobs increase permanently, basic income will most probably receive more endorsement. Whether this will be the scenario we will face in the digital economy or not remains to be seen. However, thus far it is safe to say that basic income will remain in the realms of heterodox policy in the years to come.
[i] When discussing basic income, a more reasonable approach is assessing net taxation (after-tax/after-transfer distribution). Utilising this approach, implementing a budget-neutral basic income of €490 in Finland would not lead to de facto changes in one’s taxation.
[ii] The survey is a part of the Prime Minister’s Office’s research project Finnish Work After the Transformation that contributes to drawing up the second part of the Government report on the future.
[iii] One can criticise excluding taxation in this survey, but I posit the view that this is not a fundamental problem. Firstly, budget neutral basic income models can be implemented without crucial changes in people’s net taxation. Secondly, major tax reforms raise strong emotions as the surveys by EVA and Kela indicated. This may also lead to biases. And thirdly, designing a “feasible” taxation model for a basic income requires multiple political decisions and microsimulations which were out of the scope of this study. Nevertheless, I propose that the taxation element could be included in the future surveys that will, without a doubt, be carried out after the evaluation study on the Finnish basic income experiment has been conducted in 2019.
[iv] The option “good idea” is a variable recoded from alternatives “very good idea” and “good idea” – whereas the option “bad idea” is a variable recoded from “very bad idea” and “bad idea”.