Skip to main content

Social Preferences, Well-Being and Policy

Periodic Reporting for period 3 - SOWELL (Social Preferences, Well-Being and Policy)

Reporting period: 2018-10-01 to 2020-03-31

My SOWELL projects intends to push forward the research on the theory and empirics of social preferences and well-being. The project is made up of three main parts: 1) Develop New Behavioral Measures and Analysis of Social Preferences and Well-Being, 2) Analyze the foundations of Social Preferences and Well-Being, 3) Identify policies that develop pro-social behaviors and well-being.

Part 1 – Big Data: New Behavioral Measures and Analysis of Social Preferences and Well-Being

The first part of SOWELL will develop new theoretical and empirical foundations of well-being by using large-scale behavioral measures made possible by the Big Data revolution. This approach has much to offer when it comes to expanding our understanding of social preferences and well-being: we can elicit people’s behavior by looking at their queries, and by running large-scale online behavioral surveys; we can monitor social and economic attitudes in continuous time, at a very detailed geographic level and with very large samples. We can also export experimental economics into the field with « online laboratory », e.g by running online social behavioral surveys on large-scale representative samples within societies and organizations to understand the impact of pro-social preferences in real-world orgaizations.

Part 2 – Foundations of Social Preferences and Well-Being

The second part of SOWELL exploits these large-scale behavioral measures of social preferences and well-being in order to understand their foundations. Why do social preferences vary so widely from one place to the other, and from one person to the next? How do these social preferences relate to other cognitive and social skills, and to emotional well-being? What is the role of individual life experience versus social norms and inequalities in shaping social preferences? I have written three articles on these subjects:

Part 3 – New paradigms to evaluate the effect of policies

The last part of SOWELL proposes a new paradigm for evaluating policies based on their quali/quantitative effects on social preferences and well-being. The aim is to go beyond traditional economic indicators. If social preferences and well-being are assigned high value for human development, it becomes urgent to identify the policies that foster them, and to employ appropriate indicators to capture their effects on social cooperation and happiness.

SOWELL has also created a website for the promotion of its research. Publications and description of the project are available there.
I detail below the main achievements related to the three parts of the SOWELL project.


Part 1 – Big Data: New Behavioral Measures and Analysis of Social Preferences and Well-Being

1.1 Big Data and Well-Being

We have made significant progress on this part. Some of the contributions have already been the subject of articles.

[1] “Well-Being through the Lens of the Internet”

Objective, Method, and Results
We build models to estimate well-being in the United States based on changes in the volume of Internet searches for different words, obtained from the Google Trends website. The estimated well-being series are weighted combinations of word groups that are endogenously identified to fit the weekly subjective well-being measures collected by Gallup Analytics for the United States or the biannual measures for the 50 states. Our approach combines theoretical underpinnings and statistical analysis, and the model we construct successfully estimates the out-of-sample evolution of most subjective well-being measures at a one-year horizon. Our analysis suggests that internet search data can be a complement to traditional survey data to measure and analyze the well-being of a population at high frequency and local geographic levels. We highlight some factors that are important for well-being, as we find that Internet searches associated with job search, civic participation, and healthy habits consistently predict well-being across several models, datasets and use cases during the period studied.

Publication
The paper has been submitted in Nature Human Behavior in 2017 and has received a Revise and Resubmit.


1.2 Online Laboratory and Behavioral Social Barometers

I have launched an Online Experimental Laboratory, in cooperation with the Medialab at Sciences Po and the OECD. We have proposed new measures of trust and cooperation by creating an online Trustlab on a large-scale representative sample of the French population. We have then extended this Trustlab with the OECD to run this methodology with already five other countries. We are now coordinating a large international cooperation to cover all OECD and non-OECD countries.

[2] “Trust and its Determinants: Evidence from the Trustlab Experiment”
With Fabrice Murtin, Louis Putterman, and Gianluca Grimalda

Objective, Method, and Results
This paper describes the results of an international survey on trust run in six OECD countries between November 2016 and November 2017 (France, Germany, Italy, Korea, Slovenia and the United States). Cutting-edge inference techniques drawn from behavioral science and experimental economics are combined with an extensive survey on the policy and contextual determinants of trust in others and trust in institutions. Overall, Trustlab confers some experimental validity to self-reported measures of trust in others and trust in institutions, and highlights the large scope for improving trust via policy action: 1) Self-reported measures of trust in institutions are validated experimentally, 2) Self-reported measures of trust in others capture a belief about trustworthiness (as well as altruistic preferences), whereas experimental measures also capture willingness to cooperate. Therefore, both measures are related, but should be considered rather complementary; 3) Perceptions of institutional performance strongly correlate with both trust in government and trust in others; 4) Perceived government integrity is the strongest determinant of trust in government; 5) In addition to perceived quality of institutions, social preferences and expectations, along with social capital associated activities such as neighborhood connectedness and volunteering matter for trust in others; 6) Just to highlight the scope for policy action, an increase in all significant determinants of trust in government by one standard deviation may be conducive to an increase in trust by 30 to 60%.

Publications
Article to be submitted to the American Economic Review. Results also presented in [3] “Trust, Social Progress and Well-Being”, Chapter for the handbook of the Commission on the Measurement of Economic Performance and Social Progress, Deaton-Stiglitz, 2018 forthcoming.


1.3 Social Preferences and the Organization of Production

The last part of my work in this section studies social preferences within organizations by exporting behavioral economics within firms and organizations. We have run one large experiment within the Open Source Software community and have already one paper, submitted to Management Science.

[4] “Social Exchange and the Reciprocity Roller Coaster: Evidence from the Life and Death of Virtual Teams”
With Yochai Benkler, Emeric Henry, and Jérôme Hergueux

Objective, Method, and Results
Reciprocal exchange is celebrated in the existing empirical literature as an important driver of the success of organizations. However, all of the available evidence on the impact of reciprocity on group-level success focuses on cross-sections of active, successful organizations. Using panel data on contributions to open source software virtual teams – both successful and failed – we find that groups with a larger share of reciprocators are not more successful on average. In fact, they are more likely to fail. Our working dataset consists of a sample of 5,557 projects involving 10,537 developers registered with Sourceforge – the largest online platform for developing and managing open source software at the time of our study. We show that this corresponds to an amplifying effect of reciprocity. Reciprocal behavior reinforces team dynamics and accelerates success in good periods. It also makes it harder to recover from periods of inactivity, during which reciprocators are more likely to stop contributing altogether. Managerial implications should be especially relevant for complex work environments, in which individual effort levels are difficult to observe.

Publication
This paper has been presented to the NBER Summer Institute in 2017 and conferences in Harvard, NYU, Zurich and LSE. Paper under revision in Management Science.


Part 2 – Foundations of Social Preferences and Well-Being

The second part of my project SOWELL is to use those new measures to understand the theoretical and empirical foundations of social preferences and their link with well-being. I have written two articles on this part.

2.1 Social Cognition, Social Preferences and Well-Being

[5] “Childhood Environmental Harshness Predicts Coordinated Health and Reproductive Strategies: A Cross-sectional Study Of A Nationally Representative (France) Sample”
With Lou Safra, Julie Grèzes, Nicolas Baumard, and Coralie Chevallier

Objective, Method, and Results
There is considerable variation in health and reproductive behaviors within and across human populations. Drawing on principles from Life History Theory, psychosocial acceleration theory predicts that individuals developing in harsh environments decrease their level of somatic investment and accelerate their reproductive schedule. Although there is consistent empirical support for this general prediction, most studies have focused on a few isolated life history traits and few have investigated the way in which individuals apply life strategies across reproductive and somatic domains to produce coordinated behavioral responses to their environment. In our study, we thus investigate the impact of childhood environmental harshness on both reproductive strategies and somatic investment by applying structural equation modeling to cross-sectional survey data obtained in a representative sample of the French population (n = 1015, age: 19–87 years old, both genders). This data allowed us to demonstrate that (i) inter-individual variation in somatic investment (e.g. effort in looking after health) and reproductive timing (e.g. age at first birth) can be captured by a latent fast-slow continuum, and (ii) faster strategies along this continuum are predicted by higher childhood harshness. Overall, our results support the existence of a fast-slow continuum and highlight the relevance of the life history approach for understanding variations in reproductive and health related behaviors.

Publication
Published in Evolution and Human Behavior 2017.


2.2 Micro-foundations of Social Preferences: Individual experience and Life history

[6] “Childhood harshness predicts long-lasting preference leader preferences”
With Lou Safra, Julie Grèzes, Nicolas Baumard, and Coralie Chevallier

Objective, Method, and Results
Understanding the origins of political authoritarianism is of key importance for modern democracies. Recent works in evolutionary psychology suggest that human cognitive preferences may be the output of a biological response to early stressful environments. In this paper, we hypothesized that people’s leader preferences are partly driven by early signals of harshness. We experimentally elicited children’s (Study 1) and adults’ (Study 2) political preferences using faces controlled for dominance and trustworthiness and showed that early childhood harshness has an enduring effect on adult political attitudes. Importantly, this effect was further confirmed using self-reported extreme authoritarianism (Study 2) and by the analysis of the large database of the European Value Survey (Study 3). We discuss the potential political implications of this early calibration of leader preferences.

Publication
Published in Evolution and Human Behavior 2017.


2.3 Social preferences and Social Norms

[7] “Friendship Networks, Trust and Political Opinions: A Natural Experiment among Future French Politicians”
With Yves Zenou, Nicolo Dalvit, and Quoc-Anh Do

Objective and Method
This paper shows how friendship shapes trust and political opinion in newly- formed social networks in a college that produces most of France’s top politicians. We make use of a unique natural experiment that randomly assigns new freshmen into “integration groups”. Pairs of students in the same integration group are more likely to become friends. This same- group membership thus serves as instrumental variable to estimate the effect of friendship in dyadic regressions. We find strong, robust effects of friendship on differences in beliefs after six months: becoming friends reduces the difference by half a point on a ten-point scale of political opinions. It works mostly by keeping friends from diverging, rather than pulling their opinions closer. The effect becomes insignificant for second-degree friends. The friendship effect completely dominates the effect of belonging to the same study group throughout the freshman year. The findings highlight the importance of analyzing elicited friendship data instead of using peer groups.

Publication
This paper has not been submitted yet, but has been presented at the NBER summer Institute, Harvard and Princeton seminars in 2017.


Part 3 – New paradigms to evaluate the effect of policies

The last part of my research project SOWELL proposes a new paradigm to evaluate policies based on their effects on social preferences and well-being that goes beyond traditional economic indicators. We have made progress on three dimensions.

3.1 Educational policies targeted at non-cognitive skills

[8] “The Impact of Non-Cognitive Skills Training on Academic and Non-academic Trajectories: From Childhood to Early Adulthood”
With Elizabeth Beasley, Frank Vitaro, and Richard E. Tremblay

Objective, Method, and Results
Non-cognitive skills are closely associated with adult socio-economic success. However, it is unclear whether interventions targeting those skills exclusively, rather than cognitive skills, can improve adult outcomes, and whether the window for the effective ages of intervention is wide or narrow. We show that an intervention focused on self-control and social skills at school entry changes the lifetime trajectories for children with disruptive behavior, increasing self-control and trust in adolescence, improving education achievement, and outcomes in early adulthood such as criminality, education, employment and social capital. This paper provides evidence on these questions by estimating the impact of a randomized non-cognitive skills training program at school entry for disruptive kindergarten boys from low socioeconomic environments, in particular on lifetime trajectories. The intervention consisted of a 2-year program aimed at enhancing self-control and social abilities beginning at age 7. In the spring of 1984, the Montreal Longitudinal Experimental Study (MLES) evaluated 1,037 boys at the end of kindergarten in low socio-economic areas schools. From this original sample, 250 boys were targeted for the experiment based on teacher ratings of disruptive behavior. These 250 boys were randomly assigned to either participate in the training program or to be part of the control group. Over the two-year period, 19 sessions were carried out by a team of professional childcare workers, a social worker, and a psychologist. We show that improvements in trust and self-control explain much of the impact on education and young adult outcomes, and argue that social skills are an important but neglected aspect of non-cognitive skill development. Using conservative assumptions in a simple framework, we estimate that, as a lower bound, $1 invested in this program yields about $14 in benefits over the lifetime of the participants.

Publication
This paper has been submitted to the Journal of Political Economy in 2017. Status: Revise and Resubmit. Presentations: NBER Summer Institute 2016 and seminars at Chicago, Harvard, Berkley, and NYU.

[11] “The Impact of a Large-Scale Mindset Intervention on School Outcomes: Experimental Evidence from France”
With Adrien Bouguen, Axelle Charpentier, Coralie Chevallier, and Élise Huillery

Objective, Method, and Results
This papers presents evidence on the causal effect of students’ psychological traits and mindset in school outcomes. Using a large-scale randomized experiment on a growth mindset intervention for French middle school students from disadvantaged background, we show that a light intervention focused on students’ mindset leads to positive changes in beliefs and perceptions, attitudes, and tests scores. Students benefit from a significant 6% of a standard deviation increase in their academic grades, a remarkable change given the low intensity of the program. The analysis of the mechanisms shows that the improvement of academic performance is associated with a decrease of absenteeism, a better attitude in class, as well as an increase in optimism and a decrease in perceived social determinism. Importantly, girls are the main beneficiaries: treatment girls perform 10% of a standard deviation higher than control girls, whereas for boys, the impacts are often weaker and are barely significant. Data collection is under way for the fourth year in a row.

Publication
This paper is ongoing. We presented preliminary results at the Trygfonden's Child Research Seminar, 2018.

[12] “Preventing School Violence: Evidence from a Large-Scale Field Experiment in France”
With Axelle Charpentier, Nina Guyon, and Élise Huillery.

Objective, Method, and Results
There is growing attention in policy and research to conflict and bullying at school due to their negative short and long-term effects on child’s future (e.g. Le et al., 2005; Ammermüller, 2007; Brown and Taylor, 2008), but little experimental evidence that current interventions reduce school conflicts. We use data from a large-scale field experiment in primary and middle schools in France (10,000 students) to evaluate the impact of social mediation involving the regular presence of a pro-social adult (social mediator) at school. This study also examines the characteristics of the treated students and the mediators to conclude which of these cau
Until the end period, I plan to push forward the research agenda on the three parts of the ERC. While I have already achieved most of the projects described in Part 1 and Part 2 of my ERC proposal, I still have to achieve large-scale randomized experiments to evaluate the impact of education policy on well-being and trust.


Part 1 – Big Data: New Behavioral Measures and Analysis of Social Preferences and Well-Being

In accordance with the initial ERC proposal, I will exploit the development of the Trustlab on representative samples to test the various theories about the prevalence and salience of preferences, distinguishing among self-regarding preferences, altruism (Andreoni, 1989), reciprocity motives (Dufwenberg and Kirchsteiger, 2004) and social image (Benabou and Tirole, 2006). It will be the first test in a comprehensive framework of the distribution of preferences in a real-world population.


Part 2 – Foundations of Social Preferences and Well-Being

In accordance with the initial ERC proposal, the last step of this part will consist in estimating the causal impact of inequalities on cooperation. I will exploit the Trustlab experiment on representative samples to understand the role of social norms and inequalities on cooperation and social preferences in a relevant real-world context. For this purpose, I will exploit the panel structure of the ELIPSS/CEVIPOF panel in the French context by introducing two waves of experiments. In the second wave, I will inform the participants about the unconditional average contribution, and the average contribution conditional on demographics, from the representative sample studied in the first wave. This approach will provide a unique capacity to find out how information about the real-word average behavior of others affects people’s behavior, and to identify the reference groups whose behavior affects conditional co-operation in the society.


Part 3 – New paradigms to evaluate the effect of policies

This last part is the one that includes most of the on-going projects. According to the initial proposal, I need to achieve the following experiments:

- Social Mediation to Prevent School Violence

One key policy to develop social skills at school is to fight bullying through mediation. But what characteristics should a school mediator have and what subgroup of students should be treated to maximize the effects of a mediation intervention? Previous research has failed to analyze the intrinsic characteristics of anti-bullying interventions and link these with the success or failure of such programs. The large variability of anti-bullying programs harms the comparability of papers found in literature, mainly based on non-experimental analysis. My project relies on a randomized controlled experiment that consisted of approximately 10,000 students to analyze the effects of a mediation intervention, but also to examine the characteristics of the treated students and the mediators to conclude which of these caused a positive difference in the effect of the intervention. The preliminary results confirm that male students in 6th grade, which were the subgroup with the largest index of bullying, experienced the biggest effect by the treatment. The results also revealed that agreeableness of the mediator had a significant effect on the impact of the intervention. Most students in our sample have now reached Grade 9, where they go through a national standardized exam and have to choose among different academic tracks. With the support of the French Ministry of Education, we will match at the individual level the treated and control group with their exam results and track choices. This will allow us to analyze the potential boost in cognitive skills in the long term, as well as well-being, mediated by social skills.

- “Energie Jeunes” Program

This research focuses on the role of self-discipline, perseverance and motivation in reducing school dropouts in France’s middle schools located in underprivileged areas. This growthmindset intervention is based on three sessions of 55 minutes in classrooms, every year in front of all pupils from Grade 6 to Grade 9. The preliminary evidence suggests that a light intervention focused on students’ mindset leads to positive changes in beliefs and perceptions, attitudes, and tests scores. Students benefit from a significant 6% of a standard deviation increase in their academic grades, a remarkable change given the low intensity of the program. The analysis of the mechanisms shows that the improvement of academic performance is associated with a decrease of absenteeism, a better attitude in class, as well as an increase in optimism and a decrease in perceived social determinism. Importantly, girls are the main beneficiaries: treatment girls perform 10% of a standard deviation higher than control girls, whereas for boys, the impacts are often weaker and are barely significant. Data collection is under way for the fourth year in a row. At the end of 2018, we will be able to produce 3-year impact estimates of the “Energie Jeunes” Program. In late 2019 or early 2020, we will have analyzed the final data from national databases on exam results and academic track choice.

- The impact of social diversity at school on pupils’ educational and social skills

This study is motivated by the increasing socio-economic inequalities in the French educational system and relies on a recent policy in France aimed to reduce educational inequality by encouraging a greater mixing of students from different social backgrounds within schools. First, to our knowledge, despite the importance of the issue, only few desegregation programs have been quantitatively evaluated in the United States and no program has been evaluated in France. This research project will be the first one to provide quantitative evidence on the impact of desegregation programs on middle school students in France using an “as-good-as-random” identification strategy. Data collection is under way for the first year. We measure students’ cognitive skills through standardized academic tests, school climate (peer violence), their attitudes towards social diversity, tolerance, perception of merit, optimism/fatalism, cooperation, and peer competition. The survey will then be repeated every year at the same period up until 2021, since the pupils’ follow-up will take place until the end of middle school.