All you need is trust! Results from the Finnish basic income experiment

Posted in: Basic income, European politics, Evidence and policymaking, Welfare and social security

Olli Kangas is a Visiting Fellow at the Institute for Policy Research (IPR), Director of Equal Society research programme at the Academy of Finland, and Professor of Practice at the Department of Social Research, University of Turku.

Mikko Niemelä is a Professor of Sociology at the University of Turku. 

Minna Ylikännö is Head of the Research Team at Kela, the Social Insurance Institution of Finland.

In 2017-2018, Finland was the first country in the world to implement a nation-wide basic income (BI) experiment mandated by the government and enforced by law. From the 175,000 people who were receiving basic unemployment benefits (BUB) from the Finnish Social Insurance Institution (Kela), 2,000 were randomly selected to participate in the experiment. They formed the “treatment group” (TG) and received a BI, while the remaining unemployed people formed the “control group” (CG) and did not receive any BI. Participation in the TG was mandatory in order to avoid a selection bias. The level of the BI was set to €560 per month, which is equal to the net amount of the BUB in Finland. The BI was unconditional and exempted from taxation. Furthermore, it was not reduced by any income from employment.

Here, our interest is in the relationship between BI and trust. For any kind of sustainable social organisation, mutual trust is a prerequisite. Societies with low levels of generalised trust in fellow citizens and in institutions are bound to struggle with political instability and corruption. Hence, when reforming social institutions, and when experimenting with new social models, an emphasis should be placed on how to increase trust and, in a Putnamian sense, to get people ‘to bowl together’ (Putnam 2000).

All you need is trust

Had John Lennon and Paul McCartney been sociologists, the title of their most famous and most referenced article would have been “All You Need Is Trust”. Indeed, trust is something we need in our everyday lives. We could not interact with others in any positive or productive manner if we did not trust them. And similarly, we could not properly interact with public institutions and authorities.

In academic literature, we separate two different forms of trust. Trust in fellow citizens – generalised trust – is one important aspect of trust. The second is trust in institutions. Whether people trust institutions or not has an effect on the very functioning of societies. We should never ignore the importance of trust when assessing existing national institutions or the pursuit of new policies.

Perhaps the most well-known prophet of trust is Francis Fukuyama (1995, 2011), who, in a number of best-selling books, has emphasised the crucial role of trust in creating prosperous and economically sustainable societies. According to him, the lack of mutual trust between fellow citizens and institutions has direct consequences for the nation's social order and economic trajectory. He further argues that “a nation’s well-being, as well as its ability to compete, is conditioned by a single pervasive cultural characteristic: the level of trust inherent in the society” (Fukuyama 1995, 7).

In sum, trust between fellow citizens and trust in social institutions are the glue that keeps societies together and the oil that lubricates human and economic transactions. It creates social capital. And instead of bowling alone, people interact with each other without being in constant fear that something bad may happen to them.

Against this theoretical background, one important question in evaluating the BI experiment is whether BI increases generalised or institutional trust.

Basic income and trust: research questions, data and methods

In order to study the relationship between BI and trust, we ask:

  1. How do the treatment group (BI receivers) and the control group differ in their levels of trust in the social security system?
  2. How do the BI receivers and the control group differ in their levels of trust in other societal institutions, including politics, the judicial system and the police?
  3. How do the BI receivers and the control group differ in their levels of generalised trust?

We utilised a survey targeted at the TG and 5,000 randomly selected people from the CG. The survey was conducted in October–December 2018 and included a number of questions about trust as well as questions about social and financial well-being, subjective health, job-search activity, employment and attitudes towards basic income. The response rates for both groups were low, however not exceptionally low compared to other recent survey studies. Hence, the results should be treated with care and considered indicative rather than confirmatory. In subsequent regression models, we ‘correct’ for possible biases by controlling for background factors included in the survey.

The Finnish basic income experiment was a randomised controlled trial (RCT) with an identical TG and CG at the beginning of the experiment. The same conditions apply to the survey, although we had to limit the number of people in the survey CG. In principle, any observed differences between the TG and the CG in the data can be traced back to the experiment, and a simple comparison of means establishes causality. Hence, it is possible to draw causal conclusions on the effects of the treatment. However, the low response rate cannot be overlooked when interpreting the results. Therefore, the results presented below describe indicative connections rather than give a proof of causal loops between the treatment and outcomes.

The level of trust is higher among the treatment group

Both generalised trust – trust in other people – and institutionalised trust – trust in the social security system, the legal system and politicians – were measured on a scale from 0 to 10, where the value 0 indicates total distrust and the value 10 the highest possible level of trust.

As shown in Table 1, the level of trust varies from one institution to another. Institutions enjoying high levels of trust are those for which there are no alternatives – established societal institutions such as the judicial system, the police and the social security system. Political institutions – which people can influence directly – are less trusted. In our survey, these institutions are the EU, the Parliament, political parties and the politicians.

As shown in the table, trust in both respondent groups follows the same pattern: whereas the level of trust in the police and the judicial system is higher, it is lower with regard to politicians and political parties. Trust in other people is also high, which is typical for the Nordic welfare states. Indeed, it is considered to be one of the key elements in the functioning of these bumblebee states which, against all odds, have flown and continue to fly (see e.g. Andreasson 2017; Halvorsen et al. 2016; Svendsen & Svendsen 2016). Among the whole population, the level of generalised trust is even higher, but among unemployed people the levels of trust in other people and institutions tend to be lower. They may blame institutions for their misfortune, and, more generally, their misfortune may lead to a generalised loss of trust (Honneth 1992; Hudson 2006.)

When comparing those who received BI to those who didn’t, the levels of trust in the legal system and the police are only slightly higher in the TG than in the survey CG. And after controlling for age, gender, education and the size of the household, the significance of these differences almost disappears.

However, with regard to trust in other institutions or trust in other people, the differences remain significant even after controlling for the above mentioned background variables. After receiving BI for two years, people are more likely to trust their fellow citizens and political institutions, including the social security system. The next section tries to theoretically solve the puzzle why we observe higher levels of trust among BI receivers.

How does BI possibly enhance trust?

A mundane explanation for the observed higher levels of trust among BI receivers might be that there is an unobserved heterogeneity between those in the TG and those in the CG who responded to our survey. However, there might be a more interesting story to tell, a story that is grounded in the ample theoretical literature on trust and how trust can evolve.

If we are being trusted, we are inclined to begin to trust in those trusting in us (Ostrom 1990). In the case of social policy, social security systems that are built on the principle of trust – as are typically all universal and unconditional benefits in general, and basic income in particular – generate more trust than programmes based on continued screening and means-testing. In the former case, clients can be sure that they will get what they are promised to get, whereas in the latter case they are more dependent on the mercy and decisions of the bureaucracy. Thus, the central issue here is the predictability of institutions. Predictable institutions generate procedural trust (Sztompka 1999; Tyler 2001).

Similarly, the evolution of trust can be linked to the concept of recognition (Honneth 1992; Honneth and Frazer 2003). This concept pertains to the public acknowledgement of a person's status. In the case of social policy, recognition means that the institution in question acknowledges clients as autonomous agents and takes their claims seriously. Furthermore, institutional trust is likely to evolve if people’s perception of a public authority – in our case an income transfer security system – is that the authority is concerned about its clients’ needs and taking their welfare into consideration (Tyler 2001). As a result, the clients’ perception of the treatment they receive from the authority will also be generalised to other institutions, consequently either increasing or breaking trust both in social institutions and the public authorities implementing them (Tyler and Huo 2002).

Considering the empirical results from our survey and the vast academic literature on trust, we can confidently argue that the question of the relationship between trust and BI is of the utmost relevance, and that enhancing trust should be a goal when reforming social security systems.

To conclude, institutions that trust their people create reciprocal trust, and as John Lennon and his co-author Paul McCartney might have written: “All you need is trust!”



  • Andreasson, Ulf (2017): ‘Trust – The Nordic Gold’. Nordic Council of Ministers Analysis Report. Available in
  • Fukyama, Francis (1995): The Social Virtues and the Creation of Prosperity. London: Pelican Books.
  • Fukuyama, Francis (2011): The Origins of Political Order. From Prehuman Times to French Revolution. London: Profile Books.
  • Halvorsen, Rune, Hvinden, Bjørn and Schoyen, Mi Ah (2016): ‘The Nordic Welfare Model in the Twenty-First Century: The Bumble-Bee Still Flies’. Social Policy & Society 15 (1): 57-73.
  • Honneth, Axel (1992: The Struggle for Recognition: The Moral Grammar of Social Conflicts. Cambridge: Polity Press.
  • Honneth, Axel and Frazer, Nancy (2003): Redistribution or Recognition?: A Political-Philosophical Exchange. London: Verso.
  • Hudson, John (2006): ‘Institutional Trust and Subjective Well-Being across EU’. Kyklos 59 (1): 43-62.
  • Ostrom, Elinor (1990): Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press.

Posted in: Basic income, European politics, Evidence and policymaking, Welfare and social security


  • (we won't publish this)

Write a response