World Wellbeing Panel

Understanding the effect of policy on national wellbeing

March 22, 2017

Despite dozens of years of research, we still know precious little about what policies increase national wellbeing.

  •  Professor Leonardo  Bechetti

    Professor Leonardo Bechetti

    Professor of Economics, University of Rome Tor Vergata
    Disagree
    We know that wellbeing is related to basic issues (health, income, employment) but also to issues less explored by standard economics such as quality of relational life and generativity (in the sense of J.S.Mill, that is the capacity of doing something that has good effects on other people's life or in the sense of Erikson (give life, accompany, let leave). Hence all policies in support of these points may increase national wellbeing. In addition to it wellbeing depend on the relationship between realisations and expectations (Deci and Ryan) Hence management of expectations is also crucial to it. An explosion of expectations can reduce wellbeing even in presence of good objective conditions

  •  Professor Martin  Binder

    Professor Martin Binder

    Professor of Socio-Economics at Bundeswehr University Munich
    Disagree
    I think we are actually making quite a bit of progress in understanding what policies can increase national wellbeing, at least on average. What we now should focus on are heterogeneity in the determinants of wellbeing to better understand who will profit more and who will profit less from such policies.

  •  Professor Ruut  Veenhoven

    Professor Ruut Veenhoven

    Professor of Sociology, Erasmus University Rotterdam
    Completely disagree
    We can now explain some 80% of the differences in average life-satisfaction across nations, and the data leave no doubt that people live happier in the most modern nations, that is, nations characterized by good governance, economic development, political democracy and rule of law. These macro-social conditions affect micro level life-satisfaction in various ways and a main causal path being that they allow their inhabitants more freedom, which results in a greater share of the population living a life that fits them. Trend analysis over the last 40 years shows that the ongoing process of societal modernization was accompanied by a rise in average happiness. Together with the rising life-expectancy this has resulted in an unprecedented rise in years lived happy.

  •  Doctor Anna  Alexandrova

    Doctor Anna Alexandrova

    Senior Lecturer in Philosophy of Science, University of Cambridge (UK)
    Disagree
    There are legitimate concerns about uncertainties in implementation of wellbeing policies but the existing evidence base is nonetheless in places respectable.

  •  Professor Stefano  Bartolini

    Professor Stefano Bartolini

    Professor of Economics, University of Sienna
    Completely disagree
    Decades of happiness research have provided precious knowledge about policies able to increase subjective well-being. We know that the quality of social and intimate relationships is critical to well-being and that some policies could enhance relationships. Urban policies. Cities always had relationships at the centre of their organisation. Crucial for creating relations was the common space, symbolized by the city square, where citizens of all ranks could meet. Cities were built for people; all streets were pedestrian. Then, the advent of automobiles transformed cities into places dangerous for humans. Cars have invaded common spaces, destroying the social fabric. We should create "relational" cities based on a reorganization of space and mobility. Private cars must be drastically restricted as a structural measure in order to encourage residents to use public transport. There must be a great number of squares, parks, quality pedestrian areas, sports centres, etc. These areas must criss-cross a city to form a pedestrian and cycling network. Advertising has a profoundly negative influence on people's well-being, outlook and relationships. This influence is greater in children than in adults (Schor, 2005). The amount of advertising we are exposed to should be reduced, especially that targeting children. A high tax on advertising and a ban on television ads targeting children should be seriously considered. Health care and schooling. Schooling is narrowly focused on cognitive capacities. It should be reshaped in order to provide emotional education, which it currently discourages. Health: studies show that happiness heavily affects health and longevity. Thus, healthcare systems are the end stations of distress. Unhappiness tends to turn into health problems, creating pressure on healthcare systems, whose costs grow increasingly unsustainable. We may spend too much on health care, when we could obtain better results through policies aimed at increasing well-being.

  •  Professor Jan-Emmanuel  De Neve

    Professor Jan-Emmanuel De Neve

    Associate Professor of Economics and Strategy, University of Oxford
    Completely disagree
    Despite there being much work left to be done in order to operationalize the academic research on wellbeing into actionable policy recommendations, there are a wealth of results that already have direct policy implications. Notably, of course, the negative impact of unemployment for both individuals and communities (and how this compares, for example, with the impact of inflation). The What Works Centre for Wellbeing is in the process of translating the evidence into user-friendly policy-memos.

  •  Professor Chris  Barrington-Leigh

    Professor Chris Barrington-Leigh

    Associate Professor, McGill University
    Completely disagree
    There's some truth here, in that because wellbeing (life satisfaction) is always, in principle, affected by everything, it will always be extremely difficult to isolate effects of policies, especially when they are scaled up to the national level. Beyond that, though, it seems we know an enormous amount from the literature, not with 100% confidence, but with at least as much confidence as we know the traditionally-measured effects of other macroeconomic, social, and investment policies. These policies relate to everything from unemployment insurance and labour investment policies, to child support, income redistribuiton, central bank policies, tax systems, education, and medicine. Indeed, convincing evidence from the SWB literature seems to implicate almost any policy one can think of, including the process by which we carry out management within government itself.

  •  Professor Wenceslao  Unanue

    Professor Wenceslao Unanue

    Assistant Professor, Business School, Universidad Adolfo Ibáñez
    Agree strongly
    A growing body of research has already shown what kind of policies may increase national wellbeing. Some examples are: promoting intrinsic instead ox extrinsic values (in schools, at work); using new metrics (e.g. subjective well-being indicators) for resources allocation; promoting trust and altruism; mindfulness programs; etc.

  •  Professor Mariano  Rojas

    Professor Mariano Rojas

    Professor of Economics, Universidad Popular Autónoma del Estado de Puebla
    Completely disagree
    I do believe we know important things about well-being and its drivers which are relevant for public policy and for development strategies. Among the many things we know: Economic growth has limited to nil impact on people's experience of being well while it has a high environmental cost that will impact future generations' well-being. Income in not the only driver of well-being and not necessarily the most important. Human relations are fundamental to people's well-being. Problems in the family may have a high impact on well-being. Leisure and how people use their free time are very important. Community-life matters. Work may by gratifying and job satisfaction depends on many factors beyond wages. There is heterogeneity across persons and along cultures in the importance some drivers have for people's well-being. I think these are very relevant findings from well-being research that completely change policy considerations.

  •  Professor Dan  Haybron

    Professor Dan Haybron

    Professor of Philosophy, Saint Louis University
    Completely disagree
    It is hard to know where to start! The World Happiness Report is a good example of some of what we know. Policies that reduce social capital by weakening communities, for instance, are highly corrosive. Policies that reduce major stressors, like universal healthcare and childcare, do a great deal of good. Economic growth can do a great deal of good, though perhaps not if it is too fast, but is less important than keeping unemployment low. Of course, there are many limitations to our knowledge, and we should avoid overconfidence. What works in one locale may not work in another. But we know enough that it would be irresponsible for governments to ignore information about the likely wellbeing impacts of their decisions.

  •  Professor Peter  Singer

    Professor Peter Singer

    Professor of Philosophy at Princeton
    Disagree
    Anything that reduces human suffering will increase national wellbeing. So providing a decent social welfare safety net for the poorest members of society is important. That includes universal health coverage, education, housing and at least a minimal level of income. In addition, we know that fostering cooperation, generosity, and altruism increases national wellbeing.

  •  Professor Nick  Powdthavee

    Professor Nick Powdthavee

    Professor of Behavioural Science, Warwick Business School, University of Warwick
    Disagree
    I have life satisfaction in mind here. I think "little" is a relative term; we know a lot more than ten years ago what policies increase national Wellbeing. For example, our latest work on wellbeing over the life-course highlights emotional wellbeing in childhood to be very important for adult life satisfaction compared to family economics conditions. We also know the causal effects of many things on wellbeing (e.g. Unemployment, income, retirement, etc.), which can be used to inform policies. I agree that there are a lot more that we don't know, but I wouldn't say we know very little about what improves people's wellbeing.

  •  Professor Arie  Kapteyn

    Professor Arie Kapteyn

    Professor of Economics, University of Southern California
    Agree
    this is by and large true, as the literature tends to generate fragmented and sometimes contradictory results. I do think it is pretty clear that health and economic security contribute to wellbeing and these are perfectly amenable for policy interventions

  •  Professor Richard  Easterlin

    Professor Richard Easterlin

    Professor of Economics, University of Southern California
    Completely disagree
    There is now clear evidence that welfare state policies increase well-being, as indexed by evaluative measures of SWB.

  •  Professor Felicia  Huppert

    Professor Felicia Huppert

    Emeritus Professor of Psychology, University of Cambridge
    Agree
    Although there has been a long history of research on the factors associated with wellbeing, there has been very little research on the impact of policies specifically designed to enhance wellbeing.

  •  Professor Bruno  Frey

    Professor Bruno Frey

    Visiting Professor of Economics and Wellbeing, University of Basel
    Agree
    Happiness research is incompletely connected with Political Economy.

  •  Professor Heinz  Welsch

    Professor Heinz Welsch

    Professor of Economics, University of Oldenburg
    Disagree
    There are several robust and uncontroversial findings concerning policies that increase (subjective) wellbeing. These include increasing economic stability and income security, reducing unemployment and inequality, improving work conditions and health care, reducing (air) pollution and reducing corruption. The effectiveness of such policies may depend on the initial situation. For instance, yearly working hours are low in Germany and high in the U.S. I do not think that a further reduction in Germany would create much happiness. Therefore, the issues of non-linearity/satiation points and of heterogeneity are very important. In addition, the effectiveness andfeasibility of many policies may depend on the political culture in the respective nations (e.g. attitude towards "paternalism"?).

  •  Professor Alois  Stutzer

    Professor Alois Stutzer

    Professor of Political Economics, University of Basel
    Disagree
    There are many time-proven public policies that get broad support in the population because they contribute to favorable outcomes in society. The consequences of policies for the individual well-being of different groups at the margin (more or less of some publicly provided good, a looser or stricter regulation in some area, etc.), are the tricky part. Evidence based public policy is developing in a fruitful direction.

  •  Professor Shigehiro  Oishi

    Professor Shigehiro Oishi

    Professor of Psychology at Virginia
    Agree
    Most policies are not thoroughly examined in terms of self-reported well-being before being implemented. Researchers do their best to find a natural experiment, but more systematic research with the help of policy makers is needed.

  •  Professor Jan  Delhey

    Professor Jan Delhey

    Professor of Sociology, University of Magdeburg
    Disagree
    If we conceive policies narrowly, then yes, unambiguous findings about the well-being effects of specific policies are rather scares. But more broadly, well-being research has converged about a number of over-arching policy goals which are conducive to human well-being, including: low corruption, wealth, low unemployment, (gender) equality, low inflation and so on.

  •  Professor Homa  Zarghamee

    Professor Homa Zarghamee

    Assistant Professor of Economics, Barnard College, Columbia University
    Disagree
    An excellent summary along with policy implications can be found in Helliwell (2006). Helliwell, John F. 2006. "Well-Being, Social Capital and Public Policy: What's New?" The Economic Journal.

  •  Professor John  Helliwell

    Professor John Helliwell

    Professor Emeritus of Economics, University of British Columbia
    Neither agree nor disagree
    A lot is known about the social, institutional and economic supports for national well-being. Attention is appropriately turning to develop and test alternative ways of enabling people to live better lives in their neighbourhoods and nations. There are many experiments in train, and some welcome efforts to share the results. Well-being appears to be improved most, and most reliably, where a lot of attention is given to the 'how' as well as the 'what' of policy design and delivery. In particular, mechanisms and policies that build and sustain mutual trust through shared initiatives, fuelled by generosity and friendship, seem to work best.

  •  Professor Daniel  Benjamin

    Professor Daniel Benjamin

    Associate Professor of Economics, University of Southern California
    Agree
    While there is much research correlates of happiness or life satisfaction, very little of the research convincingly establishes causal evidence.

  •  Professor Ori  Heffetz

    Professor Ori Heffetz

    Associate Professor of Economics, Cornell University and Hebrew University
    Agree strongly
    Research has provided a wealth of interesting evidence regarding the correlates of some well-being measures. To name just two examples, we have lots of evidence on how some measures change with age (an exogenous variable), and we have evidence on the correlations between well-being measures and income (an endogenous variable). But we know preciously little about causality (in the case of endogenous variables), we know preciously little about the tradeoffs individuals are willing to make across different aspects/components of well-being and across different time periods (more well-being today versus more well-being in retirement), and we know preciously little about how existing correlations might change once policy changes (something similar to the Lucas Critique in macroeconomics). And I didn't even mention issues of how to aggregate individual well-being measures into a national measure. We need more researchers to join us in trying to figure out the answers to these important questions!





In order to find out what raises national wellbeing, we need to have thousands of randomised controlled trials in all major areas of national policy.

  •  Professor Leonardo  Bechetti

    Professor Leonardo Bechetti

    Professor of Economics, University of Rome Tor Vergata
    Disagree
    We have thousands of econometric analyses but, for what I know, scarce randomised controlled trials. Much more should be done in this direction

  •  Professor Martin  Binder

    Professor Martin Binder

    Professor of Socio-Economics at Bundeswehr University Munich
    Neither agree nor disagree
    I think RCTs are a bit overhyped at the moment. Yes, they can show "what works here" (Cartwright/Hardie) but inferring that the same policy will work somewhere else can often be a bit of stretch as there are so many boundary conditions that might not be present in the somewhere else situation. RCTs are certainly useful but they are not a golden bullet...

  •  Professor Ruut  Veenhoven

    Professor Ruut Veenhoven

    Professor of Sociology, Erasmus University Rotterdam
    Disagree
    Randomized trials are useful, but possible only for a limited array of policy interventions at the meso level, not at the macro level of nations

  •  Doctor Anna  Alexandrova

    Doctor Anna Alexandrova

    Senior Lecturer in Philosophy of Science, University of Cambridge (UK)
    Completely disagree
    RCTs are neither necessary nor sufficient nor always ethical.

  •  Professor Stefano  Bartolini

    Professor Stefano Bartolini

    Professor of Economics, University of Sienna
    Agree strongly
    Randomized control trials are the basis to build evidence-based policies

  •  Professor Jan-Emmanuel  De Neve

    Professor Jan-Emmanuel De Neve

    Associate Professor of Economics and Strategy, University of Oxford
    Agree
    Carefully considered and executed RCTs across policy areas would be a major step forward in terms of obtaining causal estimated on possible wellbeing policies and advance the notion of evidence-based policy. However, much direction can already be given on the basis of the existing evidence base.

  •  Professor Chris  Barrington-Leigh

    Professor Chris Barrington-Leigh

    Associate Professor, McGill University
    Disagree
    RCTs are important and should be explored when they are feasible and ethical. However, they are particularly difficult for SWB and national-scale policy (see my comments above), and we know plenty without them. We do not need a "the science is not settled" argument for many areas of policy where the implications of SWB literature are a consistent chorus that point in a clear general direction. There are enormous opportunity costs of waiting to implement good policy, so a creative approach is needed in order to translate SWB insights into opportunities at all levels of organizations. Indeed, I believe that in the context of such complex and all-encompassing outcomes as SWB, when there are big differences across observational units (say, countries), one can start by simply adjusting course towards the policies, in general, that are in effect in the most successful units. For example, while it is incumbent on researchers to better isolate causality in accounting for the SWL success of the Denmarks of the world (and to always stay mindful that the past does not ensure future trends, and that cautious because simultaneous conflicting factors may mask causality), we should immediately look at steering our ships somewhat in the Scandinavian (say) direction for SWB-relevant factors which are being treated by policy there, simply because their blend of policies is doing noticeably better than some others in producing measurably good lives.

  •  Professor Wenceslao  Unanue

    Professor Wenceslao Unanue

    Assistant Professor, Business School, Universidad Adolfo Ibáñez
    Disagree
    Highly validated statistical processes are always useful. Especially when culture is taking into account. However, I am not sure if at this stage a??we need to have thousands of randomized controlled trials in all major areas of national policya??. More than 30 years of research has studied a??well-being for public policiesa?? (e.g. education, labor, mental health, the environment) and founded consistently what make people happier. Thus, we could start with these policies. Of course we need more research, but nowadays we have very important findings. In addition, human nature is very similar (in average). For example, Self-determination theory Deci & Ryan, 2000) has shown that our psychological needs are universal. Why would we need a??thousands of randomized controlled trialsa?? for exploring what it is important for people?

  •  Professor Mariano  Rojas

    Professor Mariano Rojas

    Professor of Economics, Universidad Popular Autónoma del Estado de Puebla
    Neither agree nor disagree
    Causality issues are very difficult to address and most researchers tend to think in terms of simple direct causality (Event A implies event B; thus, intervention in A will have an impact on B). In practice, we are dealing with complex social issues which may depend on stochastic factors and on magnifying complexity emerging from social interactions within a social context. Thus, probably we need to avoid a social-engineering approach -which assumes that we can build the society we imagine in our mind- in order to foster a co-evolutionary and institutional view so that interventions focus on setting the the stage while recognizing that people act and interact.

  •  Professor Dan  Haybron

    Professor Dan Haybron

    Professor of Philosophy, Saint Louis University
    Completely disagree
    That would be nice to have, but RCTs themselves have limitations and can only tell us so much. There's no avoiding the fact that policymakers will always face limited information about the specific, highly complex realities they face. (Elinor Ostrom did wonderful work on related issues.) But we already know a great deal about what raises national wellbeing--some of it fairly commonsensical (like, unemployment doesn't make people happy). We don't need mountains of new research to act--we've long since known enough to figure wellbeing into policy decisions. Indeed, some very effective policies haven't relied on research at all--just applying a "wellbeing lens" to policy and focusing directly on qualify of life issues. Humility is needed given our limited information, but that has always been the case, in every reasonable policy approach.

  •  Professor Peter  Singer

    Professor Peter Singer

    Professor of Philosophy at Princeton
    Disagree
    We do need some randomised controlled trials, for example of basic minimum income schemes, of the kind that Give Directly is now doing in Kenya. But "thousands" seems an exaggeration.

  •  Professor Nick  Powdthavee

    Professor Nick Powdthavee

    Professor of Behavioural Science, Warwick Business School, University of Warwick
    Disagree
    Whilst RCT is important to ur understanding of causal effects of X on Y, it also has its drawbacks -- as pointed out by Angus Deaton. It is also not the golden bullet that answers every question on causality. (One of the most important non-RCT findings is that smoking increases the risk of cancer). I believe that, in order to find out what raises national Wellbeing, a mixture of RCT, large-scaled secondary data anaylsis, and sound economic theories are more or less equally important as tools.

  •  Professor Arie  Kapteyn

    Professor Arie Kapteyn

    Professor of Economics, University of Southern California
    Completely disagree
    The great risk of doing that is produce endless cute results with very little hope for external validity

  •  Professor Richard  Easterlin

    Professor Richard Easterlin

    Professor of Economics, University of Southern California
    Disagree
    We already have a fair amount of evidence. This would cost more than it would yield.

  •  Professor Felicia  Huppert

    Professor Felicia Huppert

    Emeritus Professor of Psychology, University of Cambridge
    Disagree
    In order to find out what raises national wellbeing, we need to use sound, scientific methods, including both qualitative and quantitative assessment. These include good pilot testing of both potential programs arising from policy recommendations and the measures used to evaluate wellbeing outcomes, which recognise the multidimensionality of wellbeing. The next stage would be the adoption of large scale, randomised controlled trials of policies which aim to enhance wellbeing and their detailed evaluation, using state of the art psychometric analysis and examining outcomes across a wide range of sociodemographic variables. Adopting a high quality methodology of this kind would avoid the need to undertake thousands of separate trials.

  •  Professor Bruno  Frey

    Professor Bruno Frey

    Visiting Professor of Economics and Wellbeing, University of Basel
    Disagree
    It would take far too long for being useful for policy purposes.

  •  Professor Heinz  Welsch

    Professor Heinz Welsch

    Professor of Economics, University of Oldenburg
    Neither agree nor disagree
    I do not expect that randomised controlled trials will lead to refutation of the existing evidence, in particular if the evidence is based on (quasi) natural experiments. In addition, randomised controlled trials raise ethical issues.

  •  Professor Alois  Stutzer

    Professor Alois Stutzer

    Professor of Political Economics, University of Basel
    Disagree
    RCTs are a top-down approach to learn. There is a bottom-up approach to learn about favorable policies. This latter approach builds on the productive trial and error processes that are adopted when trying out policies in a decentralized political system.

  •  Professor Shigehiro  Oishi

    Professor Shigehiro Oishi

    Professor of Psychology at Virginia
    Disagree
    Although it is ideal to conduct randomised controlled trials, it would be probably too costly to do so.

  •  Professor Jan  Delhey

    Professor Jan Delhey

    Professor of Sociology, University of Magdeburg
    Disagree
    Although randomised controlled trials might be, in principle, a very sound base for drawing policy conclusion, there is other evidence as well, e.g. from social surveys. We do not need, for example, experiments to see that precarious jobs are detrimental to subjective well-being - we can conclude that from drawing a connection between employment conditions, and life satisfaction.

  •  Professor Homa  Zarghamee

    Professor Homa Zarghamee

    Assistant Professor of Economics, Barnard College, Columbia University
    Disagree
    While RCTs can be an important tool, we can glean plenty from national survey data and natural experiments. Privileging any one method of inquiry is unnecessary and potentially disaffecting.

  •  Professor John  Helliwell

    Professor John Helliwell

    Professor Emeritus of Economics, University of British Columbia
    Agree
    Yes, these are helpful, but equally important to seize on natural experiments and other benchmarking methods that permit a broader range of alternatives to be tried. First rule is to measure the well-being of all participants before, during and after the interventions, along with key mediating variables, of which various measures of trust are fundamental. And be prepared to be surprised by what you find, whether for not you have registered your initial hypotheses.

  •  Professor Daniel  Benjamin

    Professor Daniel Benjamin

    Associate Professor of Economics, University of Southern California
    Disagree
    Randomized controlled trials would be ideal scientifically, but they are not feasible or ethical for many important areas of national policy. More plausibly, I think we need well-being researchers to focus more on finding persuasive natural experiments (as in much contemporary research in economics).

  •  Professor Ori  Heffetz

    Professor Ori Heffetz

    Associate Professor of Economics, Cornell University and Hebrew University
    Agree
    I'm not sure "we need" is how I'd necessarily put it, but it would of course be nice to have experimental evidence on the effects of different policies. This is a crucial input that is badly needed (it's good to go on evidence when possible!). However, such experiments would only fill one gap, leaving many other crucial questions open (see my previous answer). For example, such experiments would not answer the fundamental question: How should national wellbeing be measured?