Research project · 2026–2030 · Copenhagen Business School
Funded by Independent Research Fund Denmark - Green Research
RIPPLE studies how sustainable behaviour spreads and how it stalls. We ask when the decisions of a few people cascade into self-reinforcing change, and when they instead meet friction, resistance, and a counter-current of their own.
The research question
How do variations in individual contagion thresholds and the presence of social friction affect the formation of positive feedback loops in the adoption of sustainable behaviours?
Model how variation in thresholds shapes collective tipping points.
Identify the factors that predict individual differences in those thresholds.
Examine how people decide under mixed social signals - encouragement and resistance alike.
A tipping point is a critical threshold where a small change triggers a large, often irreversible transformation. Thawing permafrost is the textbook case: past a point, the release of greenhouse gases becomes self-reinforcing.
Increasingly, researchers argue that social systems tip too - that shifts in norms, behaviours or technologies can catalyse rapid and widespread change. The rise of clean energy is often read this way. But the empirical evidence for social tipping points remains limited and mixed, and the concept has been criticised for flattening the complexity of human systems.
So the framework is compelling and the mechanisms are poorly understood. RIPPLE takes the mechanisms seriously: what actually has to be true of individual people, and of the ties between them, for a behavioural cascade to start - or to fail.
Theory
Behaviour spreads through networks by imitation, norm perception and peer influence. How far and how fast depends on two things: how many people have already adopted, and what it takes to move the ones who have not.
The number of others a person must see adopting before adopting themselves. Thresholds vary with values, socioeconomic position and personality, forming a population-wide distribution and that distribution decides whether change stays local or spreads.
Influence flows through close friends and family, but also through acquaintances and consequential strangers who bridge otherwise separate groups. The structure of ties determines who ever gets exposed at all.
One exposure is enough for information and low-cost behaviour. Identity-relevant action - the sustainability case - usually needs reinforcement from several independent sources before inertia gives way.
Sustainable choices are moralised, and moralised choices provoke resistance. Contagion can therefore spread refusal as readily as adoption: instead of accelerating, a transition may stall, fragment or reverse.
Positive tipping points are neither inevitable nor permanent - least of all where values and identity are at stake.
Approach
The packages are sequenced and integrated: from observation of real decisions, to experimental control, to behaviour tracked in socially embedded settings.
Behavioural data from two social trading platforms
One platform lets people pre-purchase sustainable goods and services; the other facilitates consumer lending to projects with and without sustainability goals. Both expose detailed funding histories - project specifics, timing, amounts, totals - so we can watch how earlier funders shape later decisions.
Thresholds are behavioural dispositions, hard to measure in advance and normally requiring costly repeated observation in identical contexts. Here we observe an average of 27.6 real funding decisions per person. Trajectory-based clustering lets us ask whether crossing a funding milestone (25%, 50%) triggers a surge of activity, and lets us approximate individual thresholds. Because both sustainable and non-sustainable projects sit on the same platforms, we can also test for moralised decision-making directly.
A one-month incentivised longitudinal experiment
A simulated social trading platform, built in oTree and run over four weekly sessions to keep cognitive load manageable. Each investment scenario varies risk, personal return, environmental impact and the level of funding already reached - so we can watch people trade off private gain against environmental benefit while the choices of others remain visible.
Environmental impact is operationalised through the Tree Task, whose vivid framing keeps the link between laboratory behaviour and real outcomes both meaningful and incentive-compatible. Funding levels are set using WP1's findings and the existing literature, where the few available estimates place tipping points between 20% and 40% adoption. An efficient design is constructed in NGENE. A baseline survey captures risk and conformity preferences, personal values, pro-environmental attitudes, belief in climate change and demographics.
A longitudinal panel survey on everyday sustainable behaviours, run through Statistics Denmark and linked to registry data
WP3 examines how sustainable behaviours spread and persist by comparing behaviours that differ in both prevalence and environmental impact across diet, transport, housing and other consumption domains. This allows us to test whether contagion, social ties and friction operate differently for behaviours that are widespread but low-impact versus rare but high-impact, using established mitigation-potential estimates such as Ivanova et al. (2020).
A stratified sample oversamples groups with higher adoption of selected behaviours, such as vegetarians and vegans, making both adoption and discontinuation observable. The baseline measures behaviours, perceived norms and social networks, including strong ties, their perceived support for behavioural goals and the weight participants attach to their opinions. Two annual follow-up waves track changes in behaviour, attitudes and social support. Adoption and discontinuation are analysed using survival models, while mixed-effects models account for repeated observations and individual differences.
Data are collected through Statistics Denmark, enabling survey responses to be linked to administrative registers. For behaviours captured directly in the registers - such as vehicle ownership, electricity use, heating and fuel type, and housing renovation - this provides an objective validation of self-reports. The linked data also enable network measures based on shared workplaces or educational institutions and potentially integration with population-scale social network data developed by SODAS.
Interactive
A toy model of the mechanism the project studies empirically. Each square is a person with their own contagion threshold - the number of adopting contacts they need before acting. Friction is the chance that crossing that threshold produces resistance rather than adoption, and resisters push back on their own neighbours.
| Round | Adopted | Resisting |
|---|
Illustrative only: a lattice with random long-range ties, not a fitted model or a project result. It is here to make the moving parts of the research question tangible - vary the threshold and the friction and watch the S-curve appear, flatten, or break.
Work plan
Milestones (M) and deliverables (D) as planned. Bars indicate the active period of each work package.
People
Most of the project's resources go towards the development of its early-career researchers.
Principal investigator · Copenhagen Business School
Leads WP1 and the project as a whole. Works on sustainable consumption and the behavioural side of environmental transitions, with a particular interest in what crowds fund and why.
Co-investigator · Assistant Professor, Copenhagen Business School
Leads WP2. An experimental behavioural scientist; the design, incentive structure and analysis of the longitudinal experiment are hers.
To be announced
WP3 is led by a postdoctoral researcher - the panel survey on everyday sustainable behaviours, the social-network mapping, and the survival analysis of adoption and discontinuation.
Output & dissemination
Three papers are the backbone, but they are not the whole of it. The project runs this website, an active social media presence and interactive simulations.
Success is measured against the stated deliverables and against publication in the target journals. Two further indicators matter as much: a measurable advance in the qualification of the early-career researchers on the team, and measurable outreach to society including participation in Forskningens Døgn.
Contact
For collaboration, data or media enquiries, and questions about any part of the project, write to the principal investigator.