Experimenting with Wellbeing: Basic Income, Immaterial Labour and Changing Forms of Productivity

This article is concerned with the recent (2017–2018) basic income experiment in Finland. This experiment attracted global attention, not least because of its break from the conditionalities and sanctions associated with social security payments in workfare states. This article stresses, however, that it is critical to understand how the Finnish basic income experiment was part of a broader programme of government-led reform in Finland. As well as establishing the experiment as a preferred mode of policymaking, this programme contained a range of strategies aimed at restructuring labour supply. The article shows how the basic income experiment should be understood as a behavioural intervention designed to enhance the wellbeing of unemployed populations at a time when wellbeing is emerging as a value-producing capacity.


Introduction
This article is concerned with basic income (BI) and especially with the recent (2017-2018) BI experiment in Finland. When the previous Finnish government (2015-2019) announced in 2015 that it would launch a BI experiment, it immediately drew attention from policymakers and BI advocates from around the world. For decades, BI has been a policy idea that has inspired politicians, researchers and civil society activists from across the political spectrum, with BI finding support from both 'pro-marketeers such as Hayek and Friedman and social democratic egalitarians' (Downes and Lansley, 2018a: 2). The COVID-19 pandemic has served to increase interest in BI not least because it has been located a fair and comprehensive measure to compensate for the economic hardships imposed by lockdowns and stay-in-place mandates (see e.g. Standing, 2020). While various BI models have been tested on small scales in different national contexts (see e.g. De Wispelaere, 2016; Downes and Lansley, 2018b;Widerquist et al., 2013), and there have been a flurry of trials and experiments in recent years (Standing, 2019), the BI experiment in Finland was initially seen as spearheading a paradigm shift in European welfare policy. This was especially inasmuch as it signalled that a post-productivist break from the conditionalities and sanctions associated with social security payments in workfare states was both feasible and possible .
As the details of the experiment were made clear, the enthusiasm towards the Finnish BI experiment turned to disappointment. An initial ambitious plan to run a series of experiments based on different BI models collided with budgetary, legal, institutional and political restrictions Kangas and Pulkka, 2016). As a consequence, the government settled on running one experiment with a duration of 2 years. A nationwide experiment began in early 2017 and ended late in 2018 and was carried out as a randomized controlled trial (RCT). Two thousand randomly selected registered unemployed people aged between 25 and 58 each received €560 BI per month. This income was unconditional and without means testing and precisely matched that previously received as unemployment benefits. As the experiment was rolled out, it became the subject of both national and international debate. It was criticized for not testing BI properly, but features consistent with it (Standing, 2019), as well as for a range of methodological limitations (see De . Concerns were also raised regarding the political framing of the experiment. This was especially inasmuch as Finnish politicians and policymakers tended to frame BI primarily as a tool to promote labour market reintegration (e.g. De Perkiö, 2019). Critics were, therefore, concerned that the BI experiment was ultimately an experiment with a new mode of labour market activation, albeit one operating without the conditionalities and sanctions of workfare. This article forwards a rather different understanding of the Finnish BI experiment. It suggests that it should be understood as a behavioural intervention designed to enhance the wellbeing of unemployed workers. This article stresses, however, that this enhancement should not be understood as a separate issue from matters of labour supply, since wellbeing is emerging as a productive, value-producing capacity. It stresses then that the experiment should be understood in terms of changing notions of productivity. This article also shows that in aiming to enhance the wellbeing of unemployed workers, the experiment formed part of a broader strategy of the restructuring of labour supply on the part of the Finnish government. To this end, it highlights how the Finnish BI experiment formed part of a broad programme of government-led reform in Finland that involved two linked aspects: the intensification of strategies aiming at a restructuring of a labour supply and the promotion of the experiment as a preferred mode of policymaking.
In regard to labour supply, it is important to make clear from the very outset that the government programme not only enabled the BI experiment to become a reality in Finland, but it also introduced an intensification of the conditionalities and sanctions of already existing workfare schemes as well as experiments with new modalities of conditionality and sanctioning (Adkins et al., 2019). The latter has represented a clear extension of the take-up of workfare policies in Finland at play from the 1990s onwards that have fundamentally altered the character and function of Finland's welfare state and restructured the labour supply in terms of notions of competition (e.g. Ahlqvist and Moisio, 2014;Kantola and Kananen, 2013). While at first glance it may seem paradoxical and contradictory that the government ran a BI experiment at the very same time that it was tightening and extending workfare, this article stresses how both BI and workfare must be understood as measures aimed at the enhancement of labour supply.
It is also important to make clear from the outset that Finland's BI experiment was a specific kind of policy intervention, namely, an experimental one. Experimental policy is gaining increasing traction worldwide and offers a fast track method of developing, testing and evaluating novel policy ideas (Adkins and Ylöstalo, 2018;Peck and Theodore, 2015;Triantafillou, 2017;Whitehead et al., 2018;Ylöstalo, 2020). As is well documented, experimental policies typically rest themselves on scientific methods. This includes the methods of the natural and psychological sciences and especially the methods of behavioural economics. The latter aim to shape human behaviour by subtly targeting the human conscious and unconscious via various 'nudge' strategies Leggett, 2014;Triantafillou, 2017). When the previous Finnish government developed and implemented the BI experiment, it did so in a context of an explicit commitment to renew and refresh the policymaking process, such that policy development took place in and as a 'culture of experimentation' (Prime Minister's Office [PMO], 2015: 26). While certainly Finland has carried out some policy experiments in the past, in the current juncture, the embrace of a thoroughgoing culture of experimentation must be understood to amount to the embrace of sciencebased experiments in policymaking. Indeed, the Finnish BI experiment had all the hallmarks of a science-based and especially a behavioural economics-based experiment, whereby the provision of an environment of choice was expected to produce a range of positive outcomes. This article shows, however, not only how the BI experiment was a behavioural intervention, but also how it was precisely through the techniques of behavioural economics that enhancements in wellbeing were to be affected.
This article draws on research from a larger project Social Science for the 21st Century that considers shifts in the economy-society relation in Finland. The project takes as its focus the process of economization, that is, the folding of the economy into society. It analyzes especially new modes of employment activation -the measures that are designed to encourage the unemployed and underemployed to become more active in their efforts to find work as well as improve their employability -and locates such activation measures as devices of economization. This article draws on data from this larger project and especially key policy documents regarding experimental policy and the BI experiment. These documents were produced by key national and international policy actors, including the government and the Finnish parliament, state officials, the Finnish Social Insurance Institution KELA, the Organisation for Economic Co-operation and Development (OECD), think tanks, consultants and researchers. The article also draws on parliamentary discussions regarding the BI experiment in Finland. In addition, it draws on semi-structured interviews with key policy actors involved with experimental policy as well as the BI experiment in Finland: researchers, experts, consultants and state officials (N = 16).
This article is structured as follows. In section The Basic Income Debate in Finland it outlines the long-standing debates on BI in Finland. In section Situating Basic Income: Experimental Policy it locates the BI experiment as part of a government-led reform in Finland and especially the take up of experimental policy. The following section, Behavioural Economics, elaborates the interconnections between experimental policy and behavioural economics and how the BI experiment should be understood as an experiment with the cognitive aspects of labour supply. The section Basic Income and Wellbeing shows how despite the government's touting of the experiment as a labour market reintegration measure, designers and advocates championed the experiment for its capacity to enhance wellbeing. The section highlights how in struggles over the meaning of the experiment, wellbeing and issues of labour market reintegration were positioned as oppositional aims. The penultimate section Basic Income and Changing Notions of Productivity shows, however, that in a context where wellbeing is increasingly valued as a productive capacity, by working to enhance wellbeing, the experiment must be understood as an intervention that aimed to enhance the productive capacities of unemployed. Finally, the conclusion reflects further on these findings, especially the implications for refocusing public debate on BI.

The Basic Income Debate in Finland
Finland has a long-standing public engagement with BI, and it has been on the policy agenda since the mid-1980s. Previous research Koistinen and Perkiö, 2014) has identified three waves of BI policy debates in Finland. The first wave took place in the mid-1980s. At that time, BI was seen as a means to streamline different social security systems, improve social security coverage and strengthen the social rights of all citizens by guaranteeing a minimum standard of living. The second wave took place in the late 1990s in the aftermath of a severe economic crisis. In that context, BI was predominantly understood as a mode of labour market activation, albeit one operating in a modality very different to the workfare measures that predominate today. Rather than as a mechanism to sanction unwanted behaviour, BI was understood as an intervention that could increase people's economic activity by eliminating incentive traps in social security systems. After a long silence, the third wave followed in 2007 and was activated by the Green Party following their tabling of a model of BI. Other political parties, researchers and civil society groups followed suit, with a number of models discussed and debated. These models shared certain features in common: minimum unemployment security as a reference point, financing BI through income taxation and leaving social assistance, earnings-related benefits and housing benefits untouched Koistinen and Perkiö, 2014).
Although these three waves demonstrate a long-standing public engagement, BI has never been a widely shared policy goal in Finland. Rather, the BI discussion has evolved around individual models or proposals provided by academics, activists and political parties (Koistinen and Perkiö, 2014). These models and proposals have been developed at critical moments, and in particular, they have all reflected 'a certain crisis of consciousness related to economic restructuring and high levels of unemployment' (Koistinen and Perkiö, 2014: 36). 1 Yet, while all of the proposed models in Finland have focused on the problem of unemployment, over time, they have approached this problem rather differently. In the 1980s, BI models aimed at solving the problem of unemployment by reducing labour supply (via e.g. the introduction of sabbatical leave, job-sharing and new civil society activities). In contrast, from the early 1990s onwards, BI models have focused on solving the problem of unemployment by activating people to work and increasing labour supply (Koistinen and Perkiö, 2014).
Despite the fact that BI proposals have conformed to prevailing political climates, prior to the recent experiment, there had not been a serious attempt to implement BI. One reason for this is that BI has often been supported by parties, groups and individuals that are politically weak (De Wispelaere and Noguera, 2012). Although individual supporters of BI can be found across most of the major political parties, the Greens and the Left Alliance and, to a lesser extent, the Centre Party have been the main proponents of the idea of BI. The former two parties have never established a position of power broker in Finnish politics, whereas in the Centre Party, BI has attained popularity only among individual politicians. The Social Democrats as well as the labour unions have always been resistant to BI due to their strong commitment to the ideal of full employment and work-based social security. Furthermore, political support for BI is scattered, and its advocates from different political backgrounds have not worked together to advance the idea. Former discussions of BI models have been highly individualized led by one activist or a political party with no real effort made to build a strong consensus across political parties and their factions (Koistinen and Perkiö, 2014). Even the advocates of BI have, therefore, lacked the commitment and political will required to advance it.
For these reasons, it was a surprise to many when the former centre-right coalition government (2015-2019) led by Prime Minister Juha Sipilä announced that 'a basic income pilot study will be performed' (PMO, 2015: 20). This surprise was not only because it was the first serious attempt to test a BI in Finland, but also because the Sipilä coalition government consisted of political parties that had not supported BI in the past. The largest party in the coalition, the Centre Party, had previously shown mild interest towards BI, but the other parties -the conservative National Coalition Party and the populist Finns Party -had a history of robust opposition Halmetoja et al., 2018). The most committed advocates -the Greens and the Left Alliance -were in opposition when the BI experiment was announced and implemented. Yet interestingly, the BI model that the Sipilä government proposed and tested was very similar to that proposed by the Greens in 2007 ). The following section shows, however, that rather than through a narrow lens of party politics, to understand how the BI experiment was adopted, it needs to be situated in terms of a broader programme of government reform. This programme established the experiment as a preferred mode of policymaking. Our analysis shows how it was because of this emphasis on experimental policy, rather than an interest in BI per se, that the experiment became a reality in Finland.

Situating Basic Income: Experimental Policy
In 2015 when the then new centre-right coalition government came to power, it proposed a major programme of reform via its Strategic Programme, Finland, a Land of Solutions (PMO, 2015). Setting out austerity as the only way forward to address rising public debt, as well as an intensification of workfare policies (Adkins et al., 2019), the programme also announced the government's intention to run a BI experiment. The proposed trial formed one part of a key project within the programme that sought to organize service provision around customer needs (PMO, 2015: 19-20). In turn, this project supported one of the government's six strategic objectives, namely, to enhance 'wellbeing and health' (PMO, 2015: 19-20). The BI experiment was framed to be able to do so inasmuch as it would address unemployment: 'The primary goal of the basic income experiment [is] related to employment promotion. The experiment, including follow-up research, aims to find out whether basic income promotes employment' (Ministry of Social Affairs and Health, 2019). The experiment was, therefore, positioned by the government as a means of labour market activation , albeit one without the conditionalities and sanctions of workfare. This positioning of the experiment as a device of employment activation was a disappointment to a research group commissioned by the government after the announcement of experiment to investigate the potential employment effects of BI trials (see . In a research interview, a member of this group commented: The assignment from the government was that we are to find out whether basic income is good for employment or not. [. . .] Nowhere in their assignment did the government mention general wellbeing, emancipation or things like that which are very much at the forefront in the international basic income discussion. To understand how the experiment became a reality in Finland, international debates on BI are, however, not the ultimate touchstone. In this regard, the experiment's precise status as a policy experiment is tantamount. As part of its programme, the government also aimed to renew policymaking in Finland and, in particular, to introduce innovation into the policymaking process (see Adkins and Ylöstalo, 2018;Elomäki, 2019;Mykkänen, 2016;Ylöstalo, 2020). After its term, the government wanted to demonstrate that it had taken 'bold steps' in this regard (PMO, 2015: 25), and to this end introduced a 'culture of experimentation' in regard to policy along with an experimentation programme consisting of an extensive set of trials and several smaller experiments (PMO, 2015: 26). The BI experiment was one of these trials and its roll-out must, therefore, be understood in the context of the turn towards experimental policy. A member of the group who designed the framework for the culture of experimentation noted during interview: 'this [BI] is about something broader, this is about transforming the way public administration works and basic income is only one of these experiments'. Making the point even more explicit, a consultant involved in the design and roll out of the experiment commented: Without this [culture of experimentation] framework the basic income experiment would not have existed, or at least that's what I would say. It has enabled it, it was possible to take it forward once experiments were already part of the agenda.
Finland is certainly not alone in the turn towards experimental policy. Policy experiments are increasingly being used by governments and other policy actors in advanced liberal states as a method of developing, testing and evaluating novel policies to address social problems (e.g. Adkins and Ylöstalo, 2018;Jones and Whitehead, 2018;Peck and Theodore, 2015;Triantafillou, 2017;Whitehead et al., 2018). While the USA, in particular, has a long history of policy experimentation (e.g. Rogers-Dillon, 2004), what differentiates policy experiments today from their earlier instantiations is that they operate at the very frontier of policy development. In addition, they are embedded in a transnational regime of policy practice managed through expert networks, especially networks of think tanks and consultancies (Peck and Theodore, 2015). The turn to policy experimentation in Finland must, then, be understood in this context of a deepening relationality of policymaking forged by complex post-national linkages. What also differentiates the policy experiments of today from their earlier counterparts is the way in which such experiments are inflected with the principles of behavioural economics, including the preferred method of behavioural interventions, the RCT. As will be shown below, run as an RCT experiment, the Finnish BI experiment was at root a behavioural trial. As one of the consultants involved in the design of the experiment noted, 'actually, the main point is the randomized field experiment, and basic income is like a case study within that' (see also Kangas et al., 2020).

Behavioural Economics
While it must be recognized that, then, it was by virtue of its experimental status that BI became a reality in Finland, it also must be recognized that the experiment incorporated certain principles from behavioural economics. The latter is a branch of economics in which insights from the cognitive and behavioural sciences, especially cognitive psychology and neuroscience, are brought together with those of economics. Behavioural economics typically seeks to come to grips with how economic decision-making is ostensibly irrational, that is, how economic decision-making does not comply with the rationality (and especially, the utility maximization) ascribed to it by orthodox economics. Behavioural economics, therefore, seeks to understand how and why empirical reality does not correspond to orthodox economists' models of that reality. To do so, it explores how cognitive, cultural and emotional factors come into play in economic decision-making.
Developed from game theory in the 1940s (Davies, 2011), but existing largely on the periphery of mainstream of economics as well as policymaking, in recent years (and especially post the financial crisis of 2007/2008), behavioural economics has become somewhat of a darling of governments not least in the applied form of nudge economics or nudgeconomics. Against the backdrop of the apparent irrationality and inefficiency of economic decision-making, the key principle of the latter is that, via the provision of choice architectures in the environment, people's behaviour can be steered (or 'nudged') to effect more predictable (and desirable) decision outcomes. Nudgeconomics (or behavioural insight), therefore, seeks to govern aspects of human behaviour that were previously classified as either insignificant or unknowable (Pykett et al., 2017;Whitehead et al., 2018). Described by its advocates as a technique to influence behaviour and improve people's welfare without coercion, or as a libertarian paternalism (Thaler and Sunstein, 2008), nudgeconomics has informed a range of public policy developments including new approaches to the payment of fines and tax collection. As John and Stoker (2019) have documented, while in the early phases of the development of behavioural or nudge policy, the focus was on such routine transactions between state agencies and the public, more recently, behavioural policy interventions have increased in scope and range to include non-transactional regulatory policies including those focusing on retirement savings, job security and labour supply. 2 Moreover, many of these novel public policy activities aiming to shape the conduct of individuals through the deployment of the behavioural and psychological sciences have taken place in experimental form .
As the scope and range of nudge policies have expanded, so too has the influence of nudge experts in policymaking. Indeed, behavioural nudge units, teams, think tanks and consulting groups comprising behavioural economists, psychologists and data scientists are playing an increasingly prominent role in the policymaking process (e.g. Davies, 2016;John, 2014;Jones and Whitehead, 2018;Triantafillou, 2017;Whitehead et al., 2018). This includes the design, roll-out and evaluation of policy experiments and trials. Central to this rise has not only been shifts to policymaking processes such that non-governmental actors are increasingly prominent (Brown, 2015), but also a process in which economic expertise (especially the expertise of economists) is privileged in policymaking practices. Indeed, the increasing influence of behavioural approaches in the policy process should be understood as part of a broader and long-term process in which social problems along with their solutions are increasingly cast in economic terms (Elomäki, 2015;Hirschman and Berman, 2014).
In this context, it is important to register that the BI experiment in Finland was designed by a consortium of experts. This included experts and consultants from KELA (the Finnish Social Insurance Institution), the VATT Institute for Economic Research, the National Fund for Research and Development (SITRA), the Federation of Finnish Enterprises and the think tank Tänk (Kangas et al., 2016: 7). The latter explicitly advocates a behavioural approach to policy design and policy experimentation and is influenced by the UK's nudge unit, the aforementioned Behavioural Insights Team (BIT). Indeed, Finland's experiment contained all of the hallmarks of a behavioural intervention. This turned on the removal of conditionalities and sanctions attached to payments, that is, the provision of unconditional income to the unemployed participating in the trial.
As mentioned in the 'Introduction' section, like most advanced liberal states, Finland operates a complex system of conditionalities and sanctions in regard to the payment of unemployment benefits. Indeed, while previously categorized as a Nordic welfare state (Esping-Andersen, 1990), from the 1990s onwards, it became clear that this categorization no longer held, not least because of a recalibration and reimagining of the Finnish economy and labour force in terms of market efficiency and competitiveness (Kananen, 2014;Kantola and Kananen, 2013). As part of this shift, the unemployed became subject to activation and workfare regimes and, in particular, to regimes in which benefit payments became conditional on activities such as the completion of job searches, training and unpaid work placements and in which failure to participate in mandated activities led to payment sanctions. Time limits were also set on benefit payments as well as compulsions to accept paid work. In short, Finland's welfare state has transformed into a workfare state (Peck, 2001). Most recently, the Finnish state both intensified and experimented with workfarism in some novel ways (Adkins et al., 2019).
Given that the BI experiment took a sample of the registered unemployed as its sample group, the experiment effectively suspended the conditionalities and sanctions attached to the payment of unemployment benefits. The 2000 randomly selected participants in the trial selected by KELA received BI payments with no strings attached for a period of 2 years. Touted as 'money for nothing' by the world's press (see e.g. Henley, 2018), even if they found employment during this period, participants in the trial continued to receive payments. In suspending the conditionalities of workfarism, and in line with the principles of behavioural economics, the BI experiment, therefore, ostensibly aimed to trial how the removal of the demands and commands of workfarism may have a positive effect on labour supply. Thus, in setting out the objectives of the experiment, KELA stated on its official website that the experiment was intended to 'reduce the amount of work involved in seeking financial assistance and to free up time and resources for other activities such as working or seeking employment' (KELA, 2019).
The BI experiment sought, then, to provide the kind of choice architecture beloved of advocates of nudging. In particular, the experiment was designed to provoke favourable, non-coerced employment decisions by releasing participants in the experiment from the burdens of being unemployed and by affording a set of choices previously not available to them. The burdens from which participants in the experiment were released were, however, not simply conceived as bureaucratic and managerial, but also as cognitive. A consultant interviewed from a think tank, for example, discussed the shift afforded by the BI experiment in the following terms: If you think about labour policy, then at least previously it was so that, I'm not terribly familiar with the current practices, and is this still true that you have to notify whether you've been at work or unemployed, during some days? [Interviewer: Yes.] Apparently, you still have to notify. If you think about this from the perspective of psychology that it becomes a self-strengthening phenomenon that if you write every day ten times that 'I'm unemployed, I'm unemployed, I'm unemployed' what kind of picture does it paint of you. If you write ten times that 'I'm a winner, I'm a winner, I'm a winner', because for people, it has been noted as well to have an impact on the human mind. [. . .] Actually, what behavioural economics or social psychology or psychology can provide is a kind of sensitivity [to the fact] that how things are presented might have a surprisingly huge impact on the end result.
Here, the consultant precisely articulates how the removal of conditionalities and sanctions attached to the receipt of unemployment benefits, in this instance daily reporting requirements, can produce an improved cognitive orientation to the condition of unemployment and how that, in turn, may have positive effects in terms of employment outcomes. The BI experiment was then an experiment in steering the cognitive orientations of the unemployed with a view to producing improved feeling states. In short, it was an experiment with the cognitive aspects of labour supply. As the consultant cited above put it, 'it [was] a very psychological trial'.

Basic Income and Wellbeing
This aim of improved, or at the very least changed, feeling states was very often captured by those interviewed who were involved either directly or indirectly in the design of the experiment by use of the term wellbeing. Indeed, advocates of BI in Finland have persistently maintained that BI is primarily a wellbeing measure, with actors involved in the design of the BI experiment maintaining that wellbeing should be thoroughly embedded at all stages of design, implementation and evaluation. These actors expressed concern that while the experiment was explicitly classified by the government as a wellbeing and health initiative (PMO, 2015), during the implementation stage and across the duration of the experiment, the government, and especially politicians and policymakers, became far too focused on employment effects to the detriment of issues of wellbeing. One of the consultant interviewees, who has been involved with the design of experimental policy, for example, expressed frustration because in his view, the experiment had not been 'properly understood' with 'important political figures [ . . .] talking about it in the wrong way'. Another interviewee, a researcher involved with the experiment, stressed: Basic income is not about employment but about liberating people and emancipation in every way. Employment is only a secondary issue. But this is what the government wanted so that's what we had to work with. [. . .] They [the government] are not interested in the fact that people are doing better, people are feeling more healthy, people are less stressed, less financial troubles, worries, distress, it's just that well, did they get a job?
These concerns have been echoed in national and international debates on BI. In these debates and discussions, and because of its emphasis on employment outcomes, the Finnish government has been critiqued for positioning the experiment as an activation trial, that is, as an experiment designed to increase labour supply via incentives (e.g. De .
Despite the government's focus on employment outcomes, a number of designers of the experiment were always committed to the idea that BI was a tool for enhancing wellbeing. This commitment was made explicit prior to the commencement of the experiment: Based on previous basic income experiments it can be expected that basic income effects not only the supply of labour but also wellbeing more generally. [. . .] Poverty and irregular income increase the risk of mental illnesses and other health problems. Basic income enables social activation which can increase wellbeing. (Kangas et al., 2016: 37-38, translated from Finnish) This commitment took concrete form during the experiment when KELA made clear its intention to measure the wellbeing effects of the trial. Interviewed during the period when the experiment was ongoing, a KELA researcher involved with monitoring the experiment discussed how his research group was collecting and analysing data not only on how BI impacts the employment status but also wellbeing. This included registered data about the participants' state of health, such as 'which prescription drugs they use and what medical diagnoses they may have'. The aim, the researcher explained, was to compare data from the 'treatment group', that is, from those in receipt of basic income, with data from the 'control group', that is, from those who were also unemployed but in receipt of regular unemployment benefits. In addition to the collection and analysis of registered data, the researcher explained how the KELA research group would also carry out a survey and interviews with participants in the experiment. With that data, he said they will be able to evaluate the 'broader effects to wellbeing, participation, political participation, societal participation, trust in institutions, how people experience their own health, do they feel healthy, leisure time and so on'. 3 The analysis of this data and an assessment of experiment have been published by the Ministry of Social Affairs and Health (Kangas et al., 2019(Kangas et al., , 2020. Here, both wellbeing and labour market supply issues are addressed. This assessment will be returned to; but before doing so, it is important to situate concerns about the effects of the BI experiment on wellbeing in a broader context, and specifically to flag that with its emphasis on wellbeing, and specifically improving the wellbeing of the unemployed, the BI experiment in Finland should be understood as part of a boom in wellness instruments, indeed as part of a turn to wellness operating across corporate and public policy worlds, and especially to wellness as an ideal (Davies, 2016). As Sointu (2005) has argued, the proliferation of wellness programmes and other wellness instruments should, however, not necessarily be taken to mean that we are experiencing objective increases in wellness (however measured or understood), but rather that wellness has become an increasingly valued aspect of subjectivity. In this context, the BI experiment in Finland could be understood as an intervention that is part of this very process, one that attempted (despite the interventions of politicians and public policy officials to frame the experiment in terms of issues of labour supply) to value the subjectivity of the unemployed in terms of notions of wellness rather than solely by notions of competitiveness and employment status.
This valuation has been made explicit in the aforementioned assessment of the experiment (Kangas et al., 2019(Kangas et al., , 2020. The authors reported that initial analysis of the first wave of data showed that the experiment did not have any effect on employment status: the BI recipients were no better or worse at finding employment than those in the control group. The impacts on wellbeing were, however, reported as significant. Here, data on self-perceived happiness and life satisfaction, trust in other people and in institutions in society, as well as overall confidence in the future were reported. According to the authors of the report, the data showed that: The wellbeing of the basic income recipients was clearly better than that of the control group. Those in the test group experienced significantly fewer problems related to health, stress and ability to concentrate than those in the control group. According to the results, those in the test group were also considerably more confident in their own future and their ability to influence societal issues than the control group. (Kangas et al., 2019: 30) Here then, rather than for the production of any significant employment outcomes, the experiment is valued for its wellbeing effects; and rather than as potential workers or soon to be workers, the unemployed are themselves defined in terms of notions of wellbeing. In this respect, Finland's BI experiment may well be understood to have been an experiment with unemployment and wellbeing.

Basic Income and Changing Notions of Productivity
Notwithstanding the findings of the assessment, what stands out regarding the contestation regarding the purpose and meaning of the experiment is that politicians, policymakers and those involved in the design, implementation and assessment of the experiment tended to make clear distinctions between issues of wellbeing on one hand and issues of labour supply on the other, with government officials valuing the latter and those involved in the design, implementation and assessment of the experiment valuing the former. Indeed, what stands out regarding the contestation is how the enhancement of labour supply and the enhancement of wellbeing were cast as oppositional and antithetical aims. This casting is, however, by no means specific to the internal dynamics of Finnish BI experiment and Finnish domestic politics. Thus, in broader debates, any positive effects of BI, including the enhancement of wellbeing, are cast as permanently at risk from external (and usually state-authorized) 'labourist' forces (see e.g. Downes and Lansley, 2018b).
This opposition between wellbeing and the administration and management of labour supply comes into question, however, when the changing nature of work is taken into consideration, especially how work has become more immaterial in character. As is well documented, as the economy has come to be based on the production of services and knowledge, labour has become increasingly cognitive, psychological and emotional in character. Not only do specific jobs and occupations (especially service-, caring-and knowledge-based occupations) explicitly demand non-tangible and immaterial inputs, but these aspects of work have also become valued in occupations that are ostensibly more physical and tangible in orientation (Boltanski andChiapello, 2005 [1999] ;Hardt, 1999;Hardt and Negri, 2000;Lazzarato, 1996;McRobbie, 2010;Morini, 2007). In short, as service and knowledge production has become central to the economy, and as the occupational structure has shifted to one that is focused on service-and knowledge-based jobs, the cognitive, psychological and emotional aspects of labour have become key assets in the production process that employers seek to harness, enhance and maximize. Given the value of these aspects of labour, or more precisely that these aspects of labour are value producing, the cognitive, psychological and emotional wellbeing of workers is of increasing concern to employers, not least because improvements in wellbeing amount to improvements in the productive potential of workers and budding workers alike.
It is in this context of the rise of immaterial forms of labour that the boom in wellbeing initiatives and especially workplace wellbeing initiatives must be placed. Indeed, Davies (2011) has mapped how as work has become more immaterial in character, a wellbeing policy paradigm has emerged, one fuelled by an alliance of economic policymakers and health professions. This alliance has generated 'a new consensus, in which the psychological and "immaterial" aspects of work . . . is what requires governing and optimizing, even for traditional manual labour' (Davies, 2011: 65, see also Davies, 2016). As Davies makes it clear, one key set of actors finding favour in this alliance has been behavioural economists. Here, the touted abilities and capacities of behavioural economics and especially of nudge economics to improve cognitive orientation, health and wellbeing through the provision of architectures of choice are positioned not simply as benign mechanisms that operate to measure and enhance wellbeing, but as techniques that promise to enhance productivity. Operating in a context where maximization of the immaterial aspects of labour is being sought, and through their experimental and RCT techniques, behavioural economics is, in other words, positioned as promising to provision of a set of interventions that will enhance economic capacity.
In an environment where immaterial labour is increasingly valued and pursued, and in the context of the rise of a wellbeing regime, it must then be understood that far from standing outside of issues of labour supply and labour productivity (or indeed, as external to the market), wellbeing is a desired productive capacity and an economic resource. It is precisely against this background that the Finnish experiment in UBI should be situated and understood. As an intervention that attempted to both maximize and measure the wellbeing of unemployed workers via the enrolment of the techniques of behavioural economics, the experiment was one that actively attempted to modulate the productive capacities of potential workers and hence enhance employability in a regime where psychological and emotional wellbeing operates as a labour market resource. The Finnish experiment in BI was then a trial in defining and recalibrating labour supply in terms of notions of health and wellbeing. Here the good potential worker was defined not so much in terms of notions of activation, competitiveness and employment status, but in terms of psychological, cognitive and emotional wellness. Commentators on the Finnish BI experiment have suggested that it was paradoxical that the experiment ran at the exact same time when the sitting government was 'pushing hard for workfare schemes and tighter benefit conditionality' (Lehto, 2018: 168), that is, when it was tightening the regime of incentivizing for work. But once it is understood that wellbeing is a productive capacity and that the BI experiment precisely aimed to enhance wellbeing, the timing of the experiment and its existence alongside conventional activation and workfare measures appear less paradoxical and contradictory. Just as activation and workfare regimes seek to modulate and recalibrate labour supply, so too did the Finnish BI experiment, albeit in ways that were aimed at maximizing wellbeing.

Conclusion
By paying attention to the political and economic context, this article has highlighted how the recent BI experiment in Finland should be understood as an experiment in the management and restructuring of labour supply. This point has also been made in previous analyses of the Finnish experiment (see De . But these existing analyses have focused on the government's explicit aim of promoting labour market reintegration and have set this aim in contrast to the emancipatory and wellbeing effects of BI. This article suggests, however, that the BI experiment should be seen as a tool of labour market activation precisely because of its aim to enhance wellbeing. It has shown how the BI experiment leans on the global trend towards behavioural change policies via policy experiments and techniques that rest on psychological and behavioural sciences. Within this framing, improving the wellbeing of the unemployed becomes a very technique of labour market activation. Moreover, the BI experiment should be understood as part of the process by which work is becoming more immaterial in character not least by valuing the subjectivity of the unemployed in terms of notions of wellness and by defining the productive potential of workers in those very terms. The BI experiment should then be understood as operating at the frontier of the management of labour supply, with the state opening out new territories in its efforts to restructure the labour.
Behavioural policy experiments and their subtle nudging techniques have been understood as a strategy by which the state is increasingly operating in a covert fashion . At first sight, it seems that there was nothing covert at all about the Finnish BI experiment, especially inasmuch it made global headlines since it was very first announced. Basic income in itself is also a topic of heated political debate in Finland as well as elsewhere. Although BI has advocates across political lines, political forces on the left tend to understand BI as a means of enhancing social equality and wellbeing, while those on the right tend to view it as a simpler alternative to social security and -especially in the Finnish case -as a means of increasing labour supply. In this context, it was hardly a surprise that the centre-right Finnish government chose to frame the BI experiment in terms of labour market activation. Within the debates over the Finnish BI experiment, however, one feature of the experiment that was universally ignored was its status as a policy experiment. Indeed, throughout the research materials -policy documents, interviews, parliamentary debates -that we have drawn on here, policy experiments were without exception seen as a positive development and as a functional way to renew society, so long as the experiments are carried out in 'the right way'. Yet, our analysis has shown that through experimental policies, and especially experiments with wellbeing, the state is engaged in attempts to modulate and harness the economic capacities of citizens in new ways. If political debate continues to ignore such new modes of policy intervention, the very means by which the immaterial capacities of citizens are being harnessed will lie beyond the boundaries of political deliberation.

ORCID iD
Hanna Ylöstalo https://orcid.org/0000-0001-5048-5622 Notes 1. The interest in BI during the COVID-19 crisis might be understood precisely in these terms. 2. While the idea of a psychologically oriented state is certainly not new (see e.g. Nolan, 1998;Rose, 1999), what is new in these developments in public policy is the explicit embrace of behavioural economics.
3. The efforts made by the research group to measure and value the wellbeing outcomes of the experiment and not to be only concerned with its economic effects might be located as aligned to a long tradition in economics and social policy that has sought to measure and value non-economic aspects of life and to position such measures as alternative indices of social progress. This tradition includes the work of welfare economists (see e.g. Sen, 1982) and feminist economists (see e.g. Waring, 2018). As we shall argue, however, the social and the economic are not so straightforwardly divided.