BIG DATA, BE HAPPY

Yann Algan, Sciences Po professor and dean of the School of Public Affairs, is an activist researcher. His ambitions? To see an increase in well-being, to see people cooperate more and better, and to see politicians to take a leaf out of their book! His innovative research using big data, which is devoted to these objectives, has been published in leading international research journals and has just been awarded a highly selective ERC Consolidator grant. This is a significant boost which Yann Algan counts on using to move us towards a more open society. Interview.

 

• What made you conceive of research that could make us happier?

Yann Algan: Growth is not enough to satisfy our aspirations, especially seeing it has long been out of reach. Meanwhile, well-being and cooperation have become the core values of our societies. This leads to a series of questions: what is well-being? How can we cooperate better? Which public policies are the most relevant for developing these dynamics? This line of research seemed all the more important – and possible! – with all the promise offered by the digital era. Firstly for the researcher; if properly exploited, big data gives us access to a wealth of information that we have never had before in the social sciences. As for society, there's no need to draw you a picture: social networks, collaborative platforms, new business models, etc.

• How can big data help us move towards a more humane economy?

Y. A.: My principle is very simple: to understand the keys to well-being, you have to start by knowing how to measure it. However, the measures of well-being that we have today are vastly inadequate, especially when it comes to the subjective well-being felt by citizens. Current measures are based on declarative surveys that ask your level of satisfaction on a scale of 0 to 10. But what does it mean to have a level of well-being of 4 rather than 5? How does that say anything about my most critical concerns? And if I am already at a very high or very low level on these scales, how do I indicate a change in my well-being over time? In addition, these surveys concern small samples of citizens, with a rather narrow timeframe and geographic coverage. If we really want to assess the impact of public policies in terms of their effects on subjective well-being, it is high time we suggested much more relevant measures. Part of my research project is to improve our measures of well-being and social attitudes by leveraging the millions of signs now available on the web and social networks.

• Can you give us an example?

Y. A.: In a recent study I conducted on the US **, I show for example that Google searches relating to health issues (e.g., how to treat migraines) or economic and social issues (e.g., searches concerning job search sites) can serve as indicators for subjective well-being in different US cities, and could be useful for developing health policies, employment policies, etc. In general, all searches or discussions on Google, Facebook and Twitter are indications of our concerns. We can also use them to measure how we are affected by economic, social and political shocks, and to understand how we perceive public policies or institutions. Finally, a major advantage of big data is that they can be precisely located in time and space.

• Could big data also be useful for fostering our desire to cooperate better?

Y. A.: Again, it is important to get back to basics. To cooperate better, we need to find out more about what drives us to do things collectively. Do we cooperate for altruistic reasons, in the interests of reciprocity or for our social image? How much of our economic activities are motivated by extrinsic factors (monetary incentives) and intrinsic factors (non-monetary incentives such as altruism or social image)?
Research in experimental economics and psychology tackles these questions in the laboratory, by having participants perform specific tasks or games. The laboratory setting has the advantage of being rigorous, but it remains frustrating because it says nothing about what happens in real life, at school, in companies and among citizens.
For this reason, we decided to explore the field by creating online platforms for experimental economics, which sound out those who participate in the many collaborative and online platforms: for civil society, think Craigslist or Wikipedia, and for business, think entreprise social networks such as Chatter, Yammer etc.

• Do you really think you can interest politicians in this type of research?

Y. A.: Absolutely. OECD involvement in these issues is proof of that. For instance, a group of experts** has been set up within the OECD dedicated to questions of well-being and social progress, with internationally renowned researchers including the new Nobel Laureate in economics Angus Deaton, and in which I am fortunate to participate. Furthermore, it is through the OECD that we are setting up a TrustLab on the Internet that will allow us to apply these measures to large population samples. The goal is then to exploit our data to identify the determinants that promote or hinder the aptitude to cooperate. The final step will be to see how these determinants could be integrated into public policy and how to build trust in institutions. 

• So is it with this data that you plan to bring about change in public policy?

Y. A.: Big data and digital technology on their own are not miracle solutions for developing cooperation and well-being! Obviously, initiatives must be undertaken on the ground to contribute to policy. To that end, we are envisaging several studies. One of them will be conducted in partnership with the Ministry of Education, as part of "La France s’engage"(Fr), and will aim to analyse the impact of the “Énergie Jeunes” programme (Fr). It is a programme for early school leavers looking to develop their capacity for self-discipline, tenacity, and self-confidence. Confidence and social skills are essential for a successful life, and can be developed through appropriate policies. If you’re not convinced, you should watch the presentation I did through TEDx on the School of Confidence, regarding the marshmallow effect and children's social skills. Self-confidence, trust in others and well-being are core attributes that go hand-in-hand. We therefore decided to work towards determining which policies and interventions could help develop the non-cognitive and social skills that foster these attributes, right from early childhood. And we measure their impact on the ground.
We are also planning a study that focuses on the social skills of young unemployed people trying to get back into work. Finally, we are going to look at management policies that could enhance both the economic and social performance of organizations.

• Because you also want to explore how digital technology and collaborative practices can contribute to economic development...

Y. A.: Yes, digital technology and all the interactions that it makes possible have generated a new economy that is still poorly understood. Who are Uber drivers? Does the uberisation of the economy mean more job insecurity or does it open up new opportunities, particularly for young people from disadvantaged neighbourhoods? To make headway on these questions, we plan to study the characteristics and motivations of those involved in this economy, whether through for-profit platforms like Uber or Airbn'b or through non-profit initiatives like Wikipedia or open source software.
In parallel, we will study the economic and social impact of these new platforms. How does the digital economy redefine the traditional boundaries of companies, break up organizational silos, or revolutionise consumption patterns? But finally and most importantly, how can we better exploit the extraordinary resources of these platforms that connect millions of people to promote a collaborative economy and collective intelligence? What regulations and incentives would be most effective? 

• You must have built a Silicon Valley for this project!

Y. A.: Not quite! Although we will have to run algorithms, one of the project's strengths is that it will bring together a large team featuring a combination of many disciplines. Anthropologists, cognitive scientists, economists, political scientists, psychologists, sociologists and of course computer scientists will be working in concert. At Sciences Po, we have a research lab – the Medialab – which has the invaluable expertise of being able to identify and process big data while drawing on advanced knowledge of the social sciences. The Medialab teams contributed a great deal to my research on confidence and I know it will be the same on this project. There are obviously other Sciences Po researchers involved. Other institutions are also on board. I'm thinking in particular of the ENS centre for cognitive and social psychology, the Berkman Center at Harvard ****, of Cepremap which has developed a well-being observatory with internationally renowned researchers, and finally of international institutions such as the OECD.

* An ERC Consolidator is a grant awarded by the European Research Council, which funds cutting-edge research. The sole selection criterion is scientific excellence.
** Big Data Measures of Well-Being: Evidence from a Google Well-Being Index in the United States (pdf, 1.52 MB)
*** OECD High Level Group of Experts on Well-Being and Social Progress
****Cooperation in Peer-Production Economy: Experimental Evidence from Wikipedia, Yann Algan, Yochai Benkler, et al.

Acknowledgements
I would particularly like to thank Bruno Latour who contributed a great deal to this project, as well as Paul Girard and the whole médialab team. At the Department of Economics, my research owes much to the work of Élise Huillery on the evaluation of education and employment policies to foster social skills, Emeric Henry for the study of collaborative platforms, and to Sergei Guriev, Quoc-Anh Do, Ruben Durante and Roberto Galbiati (Fr) who are doing remarkable research on these subjects.

Find out more
Yann Algan: detailed profile and CV (pdf, 170 kb)