2022 Summer Research Fellowship Topics

Empirical topics

Topic 1 will be supervised by HLI trustee, Dr Caspar Kaiser. Topics 2, 3 and 4 will be supervised by our research analysts, Joel McGuire and Samuel Dupret.

1. Linearity, neutrality, and comparability in survey responses to subjective wellbeing questions – a pilot study

We believe that subjective wellbeing (SWB), measured with life satisfaction or affect scales, provides more direct access to an individual’s wellbeing than other measures (such as measuring health or income and then weighting how much these affect wellbeing). Therefore, we are interested in how people respond to these measures and what the limits and strengths of these measures are.

In this project, you will assist Dr Caspar Kaiser in developing and running surveys with subjective wellbeing survey items (questions). Topics these items might tackle are linearity (is a change from 4/10 to 5/10 the same as a change from 7/10 to 8/10?), neutrality (which responses correspond to a point where there is neither positive or negative wellbeing?), and comparability (are answers comparable across respondents and across time?). For more about these topics, see Kaiser (2022) on comparabilityand Plant (2020) on cardinality.

Your role will involve reviewing (and potentially revising) survey items from Dr Kaiser, coming up with new survey items, implementing them in appropriate online survey software (e.g., Qualtrics), and recruiting participants online for the survey (via Prolific for example). Samuel Dupret, one of our researchers will provide assistance with running the survey. When you come to recruiting participants, this will likely be for a pilot of the final survey. This will enable you to find out about any particular flaws with the survey. It will also allow you to take some ownership of the project by actively participating in the analysis and write-up of the pilot survey’s results.

2 & 3. What can individuals and the state do to improve people’s wellbeing, and how do these findings inform policy?

The aim of these related topics is to produce literature reviews that would serve as stepping stones for HLI’s future work on a wellbeing approach to policy making.

First, we need to investigate Topic 2: What can individuals do to improve their wellbeing? Here you will be encouraged to find and explore a range of promising areas and synthesise what you find so that this is useful for a brief current understanding, but also for building future research. Namely, understanding how X might improve people’s wellbeing is useful, as well as finding where one might gather detailed data for a quantitative analysis of X. For example, what is the relative importance of exercise, engaging in social activities, commuting, or spending time in nature?

Investigating Topic 2, helps us to tackle Topic 3: How can the state improve people’s wellbeing? You will be encouraged to explore a range of promising areas and synthesise what you find. This work will be used to update the team’s current understanding of the literature,  but also for building future research. Areas of interest might include: green spaces, housing, migration, reducing unemployment, reducing pollution, or increasing democracy. Finding additional areas of interest will be part of the project.

Both projects will involve reviewing and summarising the respective literatures on these topics. The goal of these projects is to produce a list of the promising paths for people and governments to improve wellbeing, with the fellow giving their subjective ranking and a clear explanation of their reasoning.

4. How effective are psychedelic-assisted mental health treatments?  

Psychedelic-assisted treatments appear effective for people suffering from mental health issues. What would be the benefit of legalising such treatment in a country such as the United States?

This project would likely begin with reviewing and updating the 2020 report by Founders Pledge. This would include finding new studies on the efficacy of psychedelic-assisted psychotherapy, replacing subjective inputs with empirical estimates when possible, and finding more cost information. Then you will try to model the impact of legalising psychedelic therapy state or nationwide. The end result would be a report on psychedelic-assisted mental health treatments with a cost-effective analysis.

Theoretical topics

Topics 5 and 6 will be supervised by HLI’s director, Dr Michael Plant.

5. Using subjective wellbeing to inform longtermism

“Longtermism is the view that positively influencing the longterm future is a key moral priority of our time.” – MacAskill (2019). As a philosophy, longtermism has grown in influence inside the effective altruism community. What, if anything, can understanding how to better measure wellbeing and increase the quality of lives contribute to this conversation?

This project involves vetting ideas such as: would better measures of wellbeing help states or a general artificial intelligence to align with the interests of present and future humans? How important is improving human wellbeing promising from a ‘broad’ longtermist perspective?

6. How should we do worldview diversification? 

There are important philosophical issues to consider when figuring out how to do the most good. Reasonable people can disagree about what wellbeing is, how we account for the badness of death, and population ethics.

It is likely that we will remain uncertain in all of these issues. How should we response to moral uncertainty? A very popular view, but which has yet to receive any sustained scrutiny, is the  worldview diversification promoted by Open Philanthropy: splitting our resources and assigning some to each plausible view. What is the most plausible justification for this view? How plausible is it? How exactly should we diversify amongst worldviews? What practical implications would this have?

Additional topics

We are less likely to recruit a fellow to pursue these topics, but welcome anyone who would like to make a case for working on one of these projects. 

7. Questions related to the impact of drug liberalisation

We are interested in evaluating the cost-effectiveness of policy reform to liberalise drugs in high-income countries. To know the impact of liberalisation, we need to answer several questions.

  • How bad is being in prison? Does it have lifelong effects on wellbeing?
  • How bad is being exposed to crime?
  • Does legalising a drug reduce crime in the legalising-country? What about in the drug exporting country?
  • Does legalising drugs affect addiction? What is the direct and indirect effect of drug addiction on wellbeing?
  • How bad is addiction?
  • What’s the success rate for lobbying organisations of different sizes to change laws?

8. How bad is the really bad stuff for subjective wellbeing?

There are some terrible events that can strongly affect the wellbeing of humans. If we can identify those, and how they might relate to subjective wellbeing measures, we could start building cases for interventions or policy recommendations to reduce these events.

This project would involve conducting a literature review of horrible events and how much of a burden they cause to wellbeing. For example, war, famine, displacement, torture, persecution, crime, and modern slavery. Do we know anything about the impact of these on wellbeing?

The end goal is to synthesise what you find in exploring this topic to improve our current understanding, but also for building future research (e.g., understanding how X might damage people’s wellbeing is useful, as well as finding where one might gather detailed data for a quantitative analysis of X).

9. How happy were our ancestors?

Has humanity made progress? How bad would the collapse of civilization be, or what about just turning the clock back 200 years? To answer these questions, we need to estimate how happy our ancestors were, but people only started asking each other in regular surveys in the 1950s.

In this project, you will gather information about the trajectory of wellbeing, from as far as you can find, up to now. This will involve evaluating variables that we do have longer run data on to identify the best proxies for subjective wellbeing. For example, what can suicide rate or the sentiment of books tell us about our wellbeing in the past?

What about estimating wellbeing changes over thousands of years? Potentially, one could use the relationship between subjective wellbeing in the present and variables like average height, life-expectancy or frequency of violent death to predict subjective wellbeing backwards in time. Or would we be better off trying to understand the happiness of people living with less technology like the Amish or hunter-gatherers?

10. Do measures of happiness, life satisfaction, and depression give similar answers?

Sometimes we’re given an effect size in terms of one outcome, say depression, and we want to know what the effect would be in terms of another measure like life satisfaction. But we often don’t have the data for the outcome of interest.

One solution is to estimate the relative difference in effects they provide on a body of similar evidence to predict the effect on the missing data. Say you want to know the effect of psychotherapy on life satisfaction but only have the effect of psychotherapy on depression. But you know the effect of positive psychology interventions on depression and life satisfaction. It seems plausible that you can use this relative difference between depression and life satisfaction effects from the positive psychology literature to estimate the effects of psychotherapy on life satisfaction.

This project would involve searching for meta-analyses on interventions that summarise the effects using different measures of subjective wellbeing. You will collect this information in a spreadsheet to form a view on whether the effects of multiple interventions give similar answers across happiness, life satisfaction, and depression.