Experimentalist Governance for Epistemic & Political Learning
Lessons from Finland in governance, learning & the limits of experimentation.
Joanne Tay and Veronika Mohr
4/30/20196 min read
True social experiments should enable governance systems to learn, adapt and persist (Photo: Unsplash)
Finland is regularly ranked as one of the world’s best-governed and most socially progressive countries. Yet not long ago, in the 1990s, it faced a profound economic crisis triggered by a structural shift from resource‑based heavy industry to a knowledge‑driven ICT sector dominated by Nokia. The crisis hit both private and public sectors hard.
What followed was a series of bold social policy experiments. Two stand out: a trial of reduced working hours for employed public service workers, and a universal basic income (UBI) trial for the unemployed. Both were designed to address pressing economic and social problems. Both generated valuable insights. But neither fully achieved its promise.
Why? And what does this have to do with ESG?
Why Social Policy Experiments Matter
The Social dimension of ESG focuses on how organisations treat people – workers, communities, and other stakeholders. It includes fair wages, working conditions, social protection, and inclusive governance. When companies or governments run social policy experiments, they are effectively testing new ways to manage social risks and opportunities.
Experimentalist governance – which is a framework where broad goals are set centrally while local units have autonomy to implement and refine solutions – offers a promising model. But as the Finnish cases show, even well‑intentioned experiments can fall short if they fail to support two critical forms of learning:
Epistemic learning: Gaining new scientific or technical knowledge about what works.
Political learning: Shifting the preferences, commitments, and frames of key stakeholders (policymakers, unions, employers, and the public).
When both forms of learning occur, social policy can evolve. When they do not, experiments become symbolic gestures rather than true drivers of change.
Case Study I: The Reduced Working Hours Trial (1990s)
The Problem
In the early 1990s, Finland’s public services faced severe budget cuts and a hiring freeze. Demand for health and social services was rising. Staff were overworked and aging, while young qualified professionals could not find jobs.
The Experiment
Building on a long‑standing proposal by sociologist Paavo Seppänen, the government invited municipalities to test a 6+6 scheme: 12‑hour operating days staffed by two six‑hour shifts instead of traditional eight‑hour shifts. The goals were ambitious: increase employment, improve service availability, reduce employee exhaustion, and boost productivity.
Some municipalities followed a strict Randomised Controlled Trial (RCT) protocol. Others were given more flexibility to design their own schedules, as long as weekly hours averaged 30.
What Worked
Approximately 580 new employees were hired.
Employees reported reduced exhaustion and burnout – the most significant positive outcome.
Service availability improved, especially in home and dental care.
What Did Not Work
Productivity did not increase significantly – social and health services involve human interaction that resists simple productivity metrics.
Wages were partially reduced, with newcomers earning less for the same work, raising equity concerns.
After the trial, the government withdrew support and municipalities did not continue the scheme.
Learning Outcomes
Epistemic learning was substantial: researchers gained robust evidence on working time effects, employee well‑being, and service delivery trade‑offs.
Political learning was minimal. Stakeholders largely returned to their original positions. Unions opposed reduced wages. Employers saw no clear business case. The government moved on.
Yet the ideas did not die. They resurfaced years later in experiments elsewhere, including Gothenburg, Sweden, suggesting that epistemic learning can indirectly inform political learning over longer time horizons.
Case Study II: The Universal Basic Income Trial (2017–2018)
The Problem
Finland’s economy continued to struggle with productivity decline and job market disruption. The government under Prime Minister Juha Sipilä sought to modernise the social welfare system to address economic insecurity and employment disruption while ensuring fiscal sustainability.
The Experiment
A research working group under Kela, Finland’s social insurance agency, designed a UBI trial. It attracted global praise as “innovative” and “evidence‑based”. But the reality was constrained:
Duration: Less than two years.
Payment: €560 per month – well below the average wage.
Coverage: Only 2,000 unemployed individuals (less than 2% of the eligible population), excluding other relevant groups such as low‑income workers.
What Worked
Unconditionality of the payment was genuine. Unlike traditional benefits, UBI would not be interrupted if a recipient found work – potentially breaking the “unemployment trap”.
What Did Not Work
The trial was too short and too narrow to generate credible evidence on long‑term labour supply responses.
Concurrently, the government introduced an “activation model” requiring unemployed individuals to report job‑seeking efforts every three months, with benefit cuts for non‑compliance. This confounded any clean test of UBI’s effects.
Kela proposed expanding the trial in scope and duration. The government rejected the proposal, citing the need to wait for results, while simultaneously allowing the activation model to operate.
Learning Outcomes
Epistemic learning was severely compromised by poor design and confounding policies.
Political learning was negative. The media framed the rejection as proof that UBI is “overly ambitious” and “infeasible”. The government appeared to benefit politically from being seen as fiscally responsible, reinforcing an entrenched view that poverty reflects individual failure rather than structural barriers.
As a result, Finland missed a genuine opportunity to lead global understanding of UBI.
Analysis: What Went Wrong with Experimentalist Governance?
Both cases started with strong elements of experimentalist governance (EG): broad goals set, lower‑level units given some discretion, and evaluation planned.
But both faltered on four key EG dimensions:
In both cases, the governance framework was not truly designed for experimentation. The government used the experiments to navigate political pressure, not to genuinely learn. RCT‑style designs – while rigorous for epistemic learning – actively prevented iteration and local adaptation.
What This Means for the S in ESG
For investors, companies and policymakers, the Finnish experience offers practical lessons:
Social policy experiments must be designed for both epistemic and political learning.
Good data is not enough. Without stakeholder buy‑in and real political will, even successful trials will not scale.RCTs are not always the right tool for complex social systems. Rigid experimental designs can prevent iteration, local adaptation, and mid‑course corrections – all essential for learning in messy, real‑world settings.
Social risk is material risk. The S in ESG is not about charity. It is about understanding how labour markets, social protections, and working conditions affect long‑term value. Finland’s experiments – even their failures – generate data that responsible investors and companies can use to assess social stability, workforce resilience, and governance quality.
Conclusion: From Experiment to Impact
Finland remains one of the world’s most admired governance models. But even the best‑governed countries struggle to turn experimentalist principles into lasting political learning. To do better, policymakers must:
Allow local autonomy and diversity in implementation.
Create institutionalised peer review and exchange between implementing units.
Build in iterative revision of goals, measures, and procedures.
Resist the temptation to use experiments merely as political cover.
For the rest of us – investors, analysts, corporate leaders and ESG practitioners – the lesson is clear: genuine social progress for ESG requires governance systems that learn, adapt and persist.
Contact us to discuss how we can support your social governance and reporting needs.
---
References:
Anttila, T. (2005): Reduced Working Hours. Reshaping the Duration, Timing and Tempo of Work, in: Jyväskyla Studies in Education, Psychology and Social Research (258). Jyväskyla: University of Jyväskyla.
Anttila, T., Nätti, J. and Väisänen, M. (2006): The Experiments of Reduced Working Hours in Finland. Impact on work/family interaction and the importance of the sociocultural setting, in: Community, Work and Family 8 (2), pp. 187-209.
Barrett, C. B., and Carter, M. R. (2010). The power and pitfalls of experiments in development economics: Some non-random reflections, in: Applied Economic Perspectives and Policy, 32(4), pp. 515-548.
Böckermann, P., Laaksonen, S. and Vainiomäki, J. (2007): Who bears the burdens of wage cuts? Evidence from Finland during the 1990s, in: International Journal of Manpower, 28 (2) pp. 100-121.
Börzel, T. (2012): Experimentalist governance in the EU: The emperor’s new clothes?, in: Regulation & Governance 6 (3), pp. 378–384.
Burgoon, B. and Baxandall, P. (2002): Three Worlds of Working Time: Policy and Politics in Work-Time Patterns of Industrialized Countries. Paper presented at the American Political Science Association Annual 2002, available at:
http://aei.pitt.edu/9139/1/BurgoonBaxandall.pdf (20.09.2018).
Julkunen, R. and Nätti, J. (2001): The Modernization of Working Times. Flexibility and Work Sharing in Finland. Jyväskyla: SoPhi.
Kiander, J. (2004): The Evolution of the Finnish Model in the 1990s: From Depression to High Tech, in: VATT Discussion Papers (344). Helsinki: Government Institute for Economic Research.
Nätti, J. and Anttila, T. (1999): Experiments of Reduced Working Hours in Finnish Municipalities, in: Journal of Human Resource Costing & Accounting 4 (2), pp.45-61.
O'Connor, A. (2009). Poverty knowledge: Social science, social policy, and the poor in twentieth-century US history. Princeton: Princeton University Press.
Peltola, P. (1998): Working Time Reduction in Finland, in: Transfer: European Review of Labour and Research, 4 (4), pp. 729-746.
Reynolds, M. (2018, April 25). No, Finland isn't scrapping its universal basic income experiment. Available at:
https://www.wired.co.uk/article/finland-universal-basic-income-results-trial-cancelled (20.09.2018).
Sabel, C.F. and Zeitlin, J. (2012): Experimentalist Governance, in: Levi-Faur, D. (ed.): Oxford Handbook of Governance, pp. 169-183. Oxford: Oxford University Press.
Van Parijs, P. (2004). Basic income: a simple and powerful idea for the twenty-first century, in: Politics & Society, 32 (1), pp. 7-39.
Yle Uutiset (2017, December 27). New model to activate unemployed comes into effect amid rising criticism. Available at: https://yle.fi/uutiset/osasto/news/new_model_to_activate_unemployed_comes_into_effect_amid_rising_criticism/9994403 (15.10.2018).
