This session at the Rencontres Économiques d’Aix-en-Provence on July 5th focused on the role of and trust in experts in today’s world. Below are my thoughts on the issue:
The growing influence of “fake news” and the expansion of post-truth politics is evidence of a growing mistrust in scientific knowledge and evidence. But the key to understanding this loss of trust in science is to look not specifically at science itself, but at the broader context of inequalities and trust in government and institutions. Trust in governments stands at around 40% in OECD countries. This lack of confidence in governments and experts is a result of our economic models having failed a large share of people and the planet. We have long operated on a mantra of “grow first, redistribute later.” But with a small portion of the population capturing majority of the benefits of growth while the rest face the tough effects of the decoupling of wages from growth, we must recognize that this mantra does not deliver. In essence, the benefits of growth and integration have not trickled down.
The OECD has been documenting inequalities for many decades. We have found that the average disposable income of the richest 10% is 9.5 times that of the poorest 10% across the OECD, up from 7 times 25 years ago. The richest 10% in the OECD own around half of all household assets, whilst the bottom 40% own barely 3%.
Our recently released report A Broken Social Elevator? How to Promote Social Mobility reveals that social mobility is stalling: it takes 5 generations for the poorest children to reach the OECD mean.
It is no wonder because it is not just the bottom of the income distribution that is losing out, it is also the middle. Our recent report Under Pressure: Squeezed Middle Class shows that middle incomes have grown a third less in the last 30 years than the average of the top 10%, while the cost of housing is growing three times faster than the household median income across the OECD. Costs of healthcare and education are also rising above inflation, while uncertainty and precarity are on the rise.
The frustration of a growing proportion of people left behind by traditional economic models is driving populism. Populism distorts realities, ignores facts and builds compelling narratives based on false information that connects with people’s emotions. For example, the fact that climate change, which has been scientifically proven time and again, can be ignored is an example of the power of populist rhetoric.
Digital technologies are enabling the behaviour, whether conscious or unconscious, of ignoring science and evidence, and promoting instead the information that plays on our emotions and existing views and biases.
Furthermore, digital technologies have unfortunately undermined the relative power and influence of scientific journals. Social media and web-based sources are diffusing information very quickly, irrespective of whether it is grounded in peer-reviewed processes and evidence. When combined with the algorithmic curating and customising that some platforms use to deliver content and ads to consumers, this creates echo chambers.
This has led to the rise of echo chambers as humans have a tendency to look for information consistent with their existing beliefs, or to interpret information according to their beliefs.
Echo chambers can lead to views becoming more entrenched, can cut people off from the full spectrum of mainstream news and opinion, and can reinforce bias, including against legitimate scientific evidence and expertise.
Echo chambers are a growing problem. A 2016 study of around 50 000 people across 26 countries found that social media had overtaken television among 18-24 year olds as their main source of news.
The phenomenon of echo chambers also raises risks of political manipulation, which threatens democracy by fuelling populist and even extremist views.
Political campaigns use echo chambers to promote ideologies and discredit facts, sometimes illegally, as we saw with the Cambridge Analytica scandal, where the personal data of millions of people was harvested from Facebook for political advertising purposes.
Furthermore, the various elements of these processes are not always overseen by humans anymore. Rather, developments in artificial intelligence and machine learning are taking a lot of the human intuition and ethical instinct out of online services. These dangers add to calls for more transparency and accountability of AI systems. Thus, the OECD’s AI Principles, adopted this May and endorsed by G20 Leaders as the basis of non-binding G20 AI Principles, promote transparency, explainability and accountability in AI systems. We are now working on concrete guidelines to help countries implement the AI Principles.
Echo chambers also enable the dissemination of fake news, a strong indicator of the deterioration of trust in experts and powerful fuel for populist movements. A recent study published in Science Magazine analysed 126,000 news stories shared on Twitter between 2006 to 2017 by 3 million users. It found that a false story reaches 1,500 people six times faster, on average, than a true story does and that is not just because of bots. A key takeaway is that content that arouses strong emotions spreads further, faster, more deeply, and more broadly on social media. Fake news posts are crafted to appeal to its readers’ psychological desires.
This context is particularly alarming, as fake news can have a direct and harmful impact on people’s well-being. In many countries, false health claims are generating mistrust in vaccination programs, with citizens delaying or refusing to get vaccinated.
The World Health Organisation and UNICEF are sounding the alarm about a decline in immunization rates and a rise in measles cases, with 98 countries (among which France) reporting a higher number of measles cases in 2018 than in 2017.
The combined effect of all of this is undoubtedly crippling democracy. But we must also understand that the current state of democracy deserves scrutiny. It is not enough to support free elections. We must ensure that the process of electing leaders is fair, just, and informed by evidence. This includes limiting the influence of economic inequality on political inequality, ensuring that money doesn’t grant undue influence to certain groups or individuals.
Democratic leaders must also, once elected, be ready and willing to debate and look at facts and evidence when making decisions. Democracies will not deal with the challenges of the 21st century – technological change, climate change, rising inequality – without the full contribution of science. And if we do not elect leaders willing to support and engage with scientific fact and debate, we are all losing out. In the face of urgent social and economic needs, this must involve long-term investment in research and development, improving scientific education and training, promoting public engagement in science and helping direct the energies and ingenuity to our most pressing needs.
Cherry-picking the evidence, suppressing findings not consistent with a government or political agenda or censoring and truncating analysis not only undermine the legitimacy of science but also inhibit effective decision-making.
It is evident that ‘more facts’ or ‘more evidence’ are not enough to address these challenges. As a result of growing inequalities, people react increasingly to rhetoric that speaks to their concerns, fears, and emotions. And even if it might be convenient to blame technology, we need to take a look at our human nature and our social structures to find solutions.
If we want to reinstitute trust, we cannot use the same recipes’ that contributed to break it. We need to share the benefits of growth more equally, we need to preserve the environment; we need to move away from the supremacy of the economic profession and decision making, and listen more to people’s concerns and views. In this sense, populism is not to be entirely disparaged. Rather we can learn from it. Populism is proving that people are feeling left behind and are in need of leadership and public policy that works for them. In order to restore the legitimacy of evidence and experts we must build narratives that put people back at the centre of policy.
As an evidence based organization, the OECD is taking this on board. We are channeling new approaches that go beyond outdated models through our New Approaches to Economic Challenges initiative, looking at behavioural insights and the role of emotions and narratives in shaping public and political life.
We recognize that we need a multifaceted response. We have to acknowledge the importance of treating the root causes of the problem. This includes fixing the growth model, to make it more inclusive and fuel social mobility. To do so, we have developed the Framework for Policy Action on Inclusive Growth. This requires coordinated investments in the people, places and firms that have been left behind. We need to look at early childhood education and care, at healthcare, education, including digital skills, and quality jobs. We need to strengthen collective bargaining, as well as social safety nets.
We also need to restore trust through more responsive governments, by tackling the failings of democratic systems such as rising inequalities and corruption. The OECD has a wide range of integrity tools and also instruments to tackle tax evasion and avoidance. BEPS and Automatic Exchange of Information have yielded over 95 billion euros in additional revenue collected from taxpayers coming forward and disclosing formerly concealed assets and income through voluntary compliance mechanisms and other offshore investigations.
We also need to develop resilience to misinformation and disinformation. The OECD’s PISA Survey has moved beyond basic competencies (mathematics, science and reading) to assess education systems’ ability to equip students with core competences such as critical thinking, problem-solving, social and emotional skills. These are the “Global Competencies” they will need to develop their own understanding and navigate an increasingly complex world as responsible adults and citizens.
We also need to harness technology to engage people with science. Societal engagement can take place across the research process – from agenda setting to co-production of research and dissemination of scientific information. One example is leveraging public research infrastructures to provide a focus for citizen science. In the field of astronomy, for example, lay persons are helping to classify images of the night sky that are shared on line. Another example is hackathons – these are a common way of addressing software development challenges.
Luckily, trust in science has not collapsed completely. The Wellcome Global Monitor, a global survey of more than 140,000 people in of 140 countries showed that nearly three-quarters of people worldwide trust scientists: 54% at a medium level and 18% at a high level. Only 14% had a low level of trust in science. But it is impossible to ignore the negative consequences of the effect the expansion of fake news and post-truth politics are having on our democracies, societies, and personal well-being. We must work to combat these issues through investments in inclusive growth to ensure no one is left behind.