EEI research projects

This study aims to investigate the impact of climate conditions on subjective well-being (SWB) in Germany and the UK. It will merge objective weather information with existing panel data and focus on SWB instead of depression. Using a large data set covering nearly 25 years, the main research question is whether there is a relationship between local temperature changes and SWB in both countries. The study will also examine the effect of other weather conditions such as humidity and rain on SWB, as well as differences in the relationship among socio-economic and demographic groups. The analysis will use data from the German Socio-Economic Panel (GSOEP) and the UK Household Longitudinal Study (UKHLS) and merge it with weather data from over 1160 weather stations to form climate variables.

This study aimed to use cross-sectional UK Biobank data to derive an estimate of the influence of active and passive commuting modes and commuting distance on CVD-related biomarkers as measures of health outcomes. The analysis applied logistic regression to assess the risk of exhibiting individual biomarker values outside a predefined reference interval and standard OLS regression to estimate the relation between commuting practices and a composite CVD index. The study sample comprised 208,893 UK Biobank baseline survey participants aged 40 to 69 who use various modes of transport to commute to work at least once a week. The data set included these participants’ sociodemographic and health-related information, including lifestyle indicators and biological measures. The primary outcome was a shift from low to high-risk blood serum levels in 8 cardiovascular biomarkers: total cholesterol, LDL, HDL, triglycerides, apolipoprotein A and B, C-reactive protein, and lipoprotein (a). Our multiple regressions indicated a small negative association between the composite risk index for CVD biomarkers and weekly commuting distance. Although estimates for active commuting modes (cycling, walking) may admittedly be sensitive to different covariate adjustments, our specifications show them to be positively associated with select CVD biomarkers. Commuting long distances by car is negatively associated with CVD-related biomarkers, while the active commuting modes of cycling and walking are positively associated. This biomarker-based evidence, although limited, is less susceptible to residual confounding than that from distant outcomes like CVD mortality.

Pre-registration: doi.org/10.21203/rs.3.pex-2070/v1

It is widely acknowledged that a significant portion of the emissions reductions necessary to meet net zero targets must come from changes in individual behaviour. Encouraging more sustainable food consumption and reducing food waste and loss (FWL) have been identified as key ways to address climate change at the individual and household level. While the IPCC estimates that there is a significant “technical potential” to reduce emissions through changes in diets and reductions in FWL, there is a lack of knowledge about which climate solutions are best suited to achieve this potential.

This systematic review and meta-analysis aims to synthesise the existing research on demand-side interventions targeting sustainable food consumption and food waste behaviours of individuals and households. It includes studies that evaluate a wide range of policy interventions aimed at changing actual food consumption and waste behaviours, which have the potential to contribute to climate change mitigation.

The review will include studies that observe food consumption or food waste and loss (FWL) behaviours of individuals or households in any relevant food choice setting, including real-world settings (supermarkets, restaurants, cafeterias) or online experimental settings. It will include studies that examine at least one of the following interventions targeted at changing behaviour: monetary interventions, informational and/or educational interventions, behavioural interventions, command-and-control regulation. The review will include studies that have a valid comparison group as a benchmark, and measures actual behaviour or incentive-compatible choices that entail potential emissions reductions that can be categorised as “shifting” consumers towards sustainable food consumption or “avoiding” food waste and loss. The review will include studies that make use of an experimental design, quasi-experimental methods, or pre-post intervention designs and exclude descriptive, conceptual, theoretical and qualitative studies, as well as stated preference studies and simulation and other modelling studies.

The review forms part of an ‘ecosystem of reviews’, a large-scale evidence synthesis initiative seeking to provide a comprehensive analysis of household-scale interventions and their emissions reduction potential across multiple behavioural domains. The initiative, led by the Mercator Research Institute on Global Commons and Climate Change (MCC) in Berlin, will strengthen our understanding of the emissions-mitigation potential of demand-side climate change policies and contribute towards broader evidence synthesis efforts for upcoming IPCC Assessment reports. The reviews within the ecosystem utilise state-of-the-art AI-assisted screening procedures and follow a set of harmonised inclusion criteria.

The effect of herds on human behaviour has been presumed to be pervasive in wide range of contexts that is both applauded and blamed for a range of decision outcomes. Herding representing the alignment of the thoughts/behaviours of individuals in a group without centralised coordination. However, while herding is presumed to be pervasive, the extant empirical evidence remains sparse with the majority of the literature focused on the consequence of local interaction between individuals eg through information cascades. Our paper seeks to look beyond local interaction between individuals and rather look at how ‘global’ or ‘macroscopic’ herd inputs influence individual investment behaviour in a lending-based crowdfunding context. Global herd inputs characterised as aggregate-level herd behaviour at the time when an investment decision is made, that is operationalised as the frequency of investments by others (intensity) and the duration of investment activity (persistency). By modelling specific herd dynamics in crowdfunding investments, we show that the frequency of investments at the time of investment (intensity) and the duration investment activity at the time of investment (persistency) strongly influence how and when individuals employ social information, the actions of others, in their investment decision. More importantly we observe how herding context as compared to non-herd context influence the same individual. We find that both herd intensity and persistency have non-linear influences on social information uptake and subsequent investment behaviour. Demonstrating that social information uptake by investors is dependent not on the presence or absence of the herd, but rather on the dynamics of the herd wherein they are embedded. Different herd dynamics thus influence individuals differently as it varies the degree to which individuals employ social information. We anticipate these findings can provide insights into other areas of collective decision-making, like social media, by showing that herd dynamics influence the degree to which an individual employs social information to begin with.

A large amount of literature has studied how subjective well-being changes over the lifespan and discussed the existence of a U-shape pattern. This debate makes little reference to research on hedonic adaptation, which suggests that people rapidly return to baseline levels of well-being after experiencing even major life events. We aim to relate both research areas by modelling well-being over the lifespan using a Markov model that allows us to independently vary the frequency of experienced life events, their effect on well-being, and the adaptation processes following these events. Calibrating our model with pooled data from several large panel datasets (the British Household Survey and the UK Household Longitudinal Study, the German Socio-Economic Panel, the US Panel Study of Income Dynamics, the Australian Household, Income and Labor Dynamics dataset, the Korean Labor and Income Panel Study, the Russian Longitudinal Monitoring Survey, and the Swiss Household Panel ) and 2,836,307 person-year observations, we can identify noticeable heterogeneity across different age groups in these well-being dynamics. These results can help explaining different elements of the well-being-age pattern that, in summary, resemble a U-shape.

Pre-registration: doi.org/10.1257/rct.10381-1.0

This project incorporates 2 field experiments conducted in collaboration with an Italian restaurant brand in the UK, to assess the effect of carbon footprint information on consumer behaviour. The first experiment will focus on the provision of carbon footprint information on the table menu and aims to investigate the impact on customer’s food choices. Specifically, the study will look at the effect of carbon footprint information on the probability of customers choosing a “vegan swap” meal, which refers to an option that replaces a traditional meat-based dish with a vegan alternative.

The second experiment will focus on “grab-and-go” restaurants, where customers make their food choices quickly and often with less time to consider the options. In this experiment, carbon footprint labelling will be implemented on digital menu boards and ordering “kiosks”, which will give insight on the environmental impact of the dish. This experiment will be designed to understand the effectiveness of carbon footprint labelling, in comparison to providing carbon footprint information alone.

The results of this project will provide an understanding of how carbon footprint information can be effectively used to encourage customers to make more sustainable food choices. The findings will be useful for restaurant managers and policymakers looking for ways to promote more sustainable food choices.

Delivery food apps, such as Deliveroo or UberEats, have become increasingly popular in recent years, with projections suggesting that up to 100 million people will be using them across Europe by 2024 (Statista 2022). These apps have made it easier than ever before to have meals delivered directly to our homes, providing a convenient and quick option for busy individuals. However, despite their rapidly growing importance, the impact that these apps have on our diets and the environment has received little attention. Encouraging more sustainable diets on online delivery platforms is important, as it can help to reduce the environmental footprint of our food choices, while also promoting healthier eating habits.

The goal of this study is to understand the potential of using behaviour-based policies to promote environmentally friendly food choices on food delivery apps. In order to achieve this, an interactive web-platform that mimics an online takeaway food setting will be used. The platform will include 3 treatment conditions: a meat tax, a carbon footprint label, and a choice architecture intervention, and a business-as-usual control condition. The meat tax treatment will apply a surcharge on meat dishes, while the carbon footprint label treatment will provide information on the carbon footprint of each dish. The choice architecture intervention will re-order the menu so that the lowest carbon-impact dishes are presented first. The study design will allow us to compare the effectiveness of a conventional monetary intervention, information provision and a behavioural intervention in the same experimental setting, and aim to generate new understanding of the potential impact on both personal welfare and distributional effects.

To explore these questions, we will conduct a large-scale incentive-compatible online Randomised Controlled Trial (RCT) with a sample representative of UK adult internet users. The project will be conducted in collaboration with the Behavioural Insights Team (BIT) and will be hosted on Predictiv, BIT’s policy testing lab. For this experiment, we will use a simulated online takeaway app (‘Take a BITe’), which mimics a real-world online food delivery setting. Data collection will take place in January 2022.

Pre-register online

This field experiment will be conducted in collaboration with 13 college cafeterias to test the effectiveness of sustainable food subsidies, carbon footprint labelling and the combination of both instruments on food choices. The introduction of both interventions will be accompanied by an information campaign to raise awareness about the interventions amongst students. The study will measure the impact of these interventions on the proportion of sustainable food choices among the university students who frequent the participating cafeterias and estimate potential emissions savings.

The food subsidies will be applied to all vegan meals sold in the respective intervention cafeterias, as this food group has the lowest emissions intensity. The carbon footprint labels will provide information on the carbon footprint of each dish and will be displayed next to the dish in the servery. The study will also measure the acceptability of these interventions among the university community, through surveys and interviews with a sample of participants.

The results of this study will provide valuable insights into the effectiveness of food subsidies and carbon footprint labels in promoting sustainable food choices among university students. Additionally, the study will examine the potential of combining food subsidies with carbon footprint labels to create a synergistic effect on sustainable food choices, which could help inform policymakers and university administrators looking to promote more sustainable food systems on campus and in other food choice settings.

The project examines the connection between environmental attitudes, socioeconomic status, and demand for CO2 (carbon emissions/carbon footprint). While previous research has explored the individual relationships between socioeconomic status and CO2 demand, CO2 demand and environmental psychological traits, and socioeconomic status and environmental attitudes, there is limited research on the interaction of these three factors. The hypothesis is that these 3 concepts have a systematic relationship, which can be measured as changes in “carbon elasticity of income”, “carbon elasticity of different psychological characteristics”, and “income elasticity of different psychological characteristics”. A large-scale, cross-cultural online survey will be conducted to gather data and provide a comprehensive understanding of these individual factors. The study is exploratory and will simulate the long-term effects of changes in income and psychological characteristics on CO2 demand using a simple formal framework. A better understanding of these relationships can inform environmental policies and prevent unintended consequences.

This experimental lab study examines energy-saving behaviour under different behavioural interventions. We analyse the impact of interventions presented in the form of environmental information and monetary savings in a modified public good game. Participants in the experiment will make several energy-saving decisions. Reaching a Green Neighbourhood, for example, can be considered a public good that benefits everyone when full cooperation is achieved. We also aim to shed light on the impact of ambiguity attitudes on pro-environmental behaviour, the expectation being that participants affected by ambiguity aversion will be less likely to reduce energy consumption.

This study provides evidence that choice overload affects both donors and donation levels positively, contrary to the choice overload hypothesis. In an online experiment, we explore patterns of donations in charitable giving (N = 3200). Participants allocated donations from their endowment to various numbers of charities within a charitable category. They were randomly allocated to choose either one or multiple charities to donate to. Faced with a lengthy list of charities and limited to choosing only one charity (1Choice treatment), participants donated slightly more than those participants facing a concise list, implying a sense of choice deprivation. The presence of a default charity reduced donations by a small amount. Conversely, when participants had the freedom to donate to multiple charities (Mchoice treatment), the donation amounts substantially increased relative to 1Choice treatment, irrespective of the length of the list of charities. This suggests that choice deprivation and limitation may have more significant consequences than choice overload. Participants who had many options were more likely to regret their choices but also more satisfied with them, a curious result that we discuss in-depth. In addition, those who had the option of a default perceived their choices as less complex. Neither regret, satisfaction, nor complexity of choice seems to mediate the relationship between our various treatments and the amount donated, indicating that these variables do not drive choice deprivation. The present research has implications for addressing charities’ reduced grantmaking capacities and economic pressures on individual donations.

The financial support for this project comes from an anonymous long-term benefactor of the University of Cambridge.

The study design is pre-registered at Center for Open Science.

This study provides evidence that biases attributed to the perception of probabilities affect cooperation levels in repeated games. In an experiment, subjects completed a prisoner’s dilemma game that continued to the next round with a fixed probability. Under the standard assumption of (constant) discounted expected utility, such a probability can be interpreted as time discounting. The presence of probability biases leads to deviations from constant discounting, as shown in Halevy (2008, AER), which then affects the evaluation of the outcomes in the game and, the hypothesis is that this affects the cooperate-defect decision of subjects. Using an incentive-compatible mechanism based on scoring rules, we quantify the direction and magnitude of subjects’ probability bias. We find that 53% of subjects are expected utility (EU) subjects; 21% of subjects reveal biases that accord with prospect theory: small probabilities are overweighted and medium to large ones are underweighted (inverse-S); 26% of subjects underestimate small probabilities and overestimate large probabilities (S-shape). The main finding is that the cooperation level is correlated with the type of biases. Specifically, for all continuation probabilities, inverse-S subjects cooperate more than EU subjects, and S-shape subjects cooperate less than EU subjects. We explain this behaviour in repeated games by adopting Halevy’s impatience index.

We are currently drafting the results and preparing them for publication in an international peer-reviewed journal.

RCT ID: AEARCTR-0009613

Consumers typically neglect durability or usage frequency considerations when respective cues are absent in the purchase context. The cost-per-wear (CPW) concept suggests evaluating a garment in terms of the hypothetical price paid for each time it is worn, which decreases the more often the garment can be worn. High-quality items inherently provide a higher number of wears such that despite their often higher total purchase price, their CPW is lower than for low-quality items. Based on the literature on unit prices, consumers should prefer the more economical, higher-quality clothing option (i.e. the item with the lowest CPW) when CPW information is provided.

Across several pre-registered experimental studies, we empirically test the CPW concept to provide an objective, quantifiable and comparable metric across different clothing options. We experimentally manipulate the presence of CPW information and vary its operationalisation as well as the type of stimuli (e.g. sweaters). We analyse relevant outcome measures (e.g. choice), potential process variables that drive the effect, and possible boundary conditions.

With this series of studies, we seek to provide robust evidence of the effectiveness of CPW as a simple, yet powerful intervention. If adequately implemented, CPW provides consumers with an objective indication of quality and a quantitative cue for the most economical option, making cheap fast fashion appear more expensive due to a higher CPW. Hence, CPW has the potential to stimulate sustainable decisions without explicitly cueing sustainability.

Urgent global actions are required to address climate change and reduce its risks for people and ecosystems. A large gap exists between the support needed and the effective actions taken. Climate mitigation is particularly challenging at the individual and collective levels since it requires fundamental changes to how people make decisions, including changes in consumption patterns and energy use. Addressing climate change requires people to sacrifice their present comfort and convenience for future gains. This is often seen as a manifestation of present bias or temporal discounting, a cognitive tendency to favour immediate rewards over delayed ones. Some researchers and journalists have argued that present bias is a natural and universal trait of human psychology, which makes climate change appear inevitable and climate action futile. However, this view neglects that humans can overcome their biases through various strategies such as mental time travel, investments in education, or creating supportive institutions. Our focus merges economics and psychology with urgency in exploring the intersection of individual beliefs and global climate action. Using 2 online surveys, we aim to test how the perception of human nature as either present biased or not affects people’s willingness to engage in pro-environmental behaviour. We will recruit participants from India and the United States, two major emitters of greenhouse gases, to examine the cross-cultural differences and similarities in this phenomenon. Our findings can inform policymakers on communicating the urgency of climate change and creating a more positive and proactive attitude towards climate change.

This project is funded by the Tony Cowling Foundation.

Behavioural insights units around the world bring together the expertise of behavioural research and policymaking to improve public policy by developing interventions to influence behaviour. However, little is currently known about how these units operate and the variety of approaches they employ. This project seeks to fill this gap by mapping and making sense of the diversity of approaches taken by different units. Additionally, we will investigate how these approaches correspond to the effectiveness of these units in achieving their intended impact. For example, some units focus on intervention development, while others prioritise capacity building and disseminating knowledge of behavioural tools. Overall, this project aims to provide valuable insights into best practices of integrating behavioural insights in organisations and public policy.

Behaviourally informed interventions (BIIs) are instruments to change behaviour that make explicit use of behavioural knowledge. While some interventions more commonly use behavioural knowledge (eg nudges), other instruments (eg taxes) can, in principle, be BIIs too.

In viewing BIIs as complex configurations, we aim to develop an extended and context-rich classification of BIIs. In addition, we aim for a co-designed, inter-, and transdisciplinary classification because BIIs are used and studied by multiple actors (eg policymakers, front-line bureaucrats, businesses) from diverse disciplinary and professional perspectives.

Decision-makers in public policy can have different views on behaviour. For instance, decision-makers can view behaviour as stable ‘like cement’ or changeable ‘like clay’. Such views on behaviour influence how decision-makers understand the social world around them and, specifically, what policy preferences they hold for changing behaviour. When behaviour change is urgently needed, more government intervention can be preferred from those viewing behaviour as hard to change (‘a lot helps a lot’; Dewies et al, 2022).

In this experimental study, we aim to activate different views on behaviour and study their influence on policy preferences.

The Behavioural Insights for a Circular Society – BEACON – project explores and tests behavioural changes towards sustainable lifestyles to support the building of a circular society. While our research focuses on urban food systems and experiments in a real world setting of a city – Copenhagen – we expect that our findings will apply to other consumption areas and systems of provision. The project is funded by the Novo Nordisk Foundation, the PI is Copenhagen Business School, and the El-Erian Institute is a collaborator.

This project explores the application of nudging as management practice to promote employee behaviour change for corporate sustainability. We conduct a field experiment in a German automotive company to test the effects of 3 nudges on pro-environmental actions in terms of electric vehicle choices of 170 employees. The nudges take the form of messages that emphasise different aspects of electric cars (emotional, sustainability and cost saving) and are applied at 2 times during the decision process: first in emails to remind employees about ordering a new car and second in pop-up notifications, appearing in the online system where employees complete their orders. We find that the timing of the nudges matters, with interventions applied in emails, but not in pop-up notifications, having significant positive effects on electric vehicle adoption. Yet, the durability of the effects is limited. Overall, the cost saving message has the longest and most powerful impact on electric car choices. This finding has implications for workplaces where employees might not yet possess strong pro-environmental believes, showing that employee sustainable behaviour can be enhanced by emphasising financial benefits.

Technology continues to shape the way we live and work. Simultaneously, the climate crisis is coming to a head and inequalities are on the rise – while Russia’s war on Ukraine has shattered political order. In light of such multiple crises, it is essential to govern technology in a way that ensures it serves solving challenges of social and environmental sustainability, rather than aggravating them.

Read more about the project and access the report.

Read our feature article about the project “Seven principles for using digital technology for sustainability”