Author: Teis Rasmussen
The Happier Lives Institute intends to research which careers will do the most to increase happiness . Here, we sketch out the current state of this research project. The document sets out: the project’s goals; a list of careers we intend to consider; thoughts on the correct evaluation methodology; some key uncertainties involved in this research.
1. Project goals
In prioritised order, the current research goals are:
We expect that significant parts of these goals can be addressed by relying on the work of 80,000 Hours (‘80k’); a non-profit focused on providing evidence-based career advice to (recent) graduates interested in having high amounts of social impact with their career.
Through our research, we will focus on identifying the careers that seem optimal for increasing world happiness ‘near-term’, i.e. the current generation, and we intend to base our estimates of expected increases in happiness on self-report data from the relevant research literature.
80,000 Hours primarily focus on the ‘long-term’, i.e. benefitting future generations, and their recommendations regarding the near-term are not explicitly or solely based on self-reported happiness scores. Thus, because of our different focus, we expect to rely on our own research in achieving goal #1. For goals #2 and #3, we expect to draw heavily on the work of 80,000 Hours: we expect their suggestions on those issues can be applied generally although we might end up proposing alterations to their processes.
For the remainder of this research agenda, we’ll focus on explaining our approach to goal #1.
2. Potentially promising careers
Here, we outline the careers we intend to investigate further and some that don’t presently seem worth looking into.
2.1 Careers currently under consideration
The list below is not meant to be exhaustive. We expect it to change as we learn more about various careers and their potential to affect happiness globally. It is organised by type of work.
Careers in research
Advocacy and policy
2.2 Careers that are currently not under consideration
The following options seem like they might be promising in their potential to increase happiness, but they were ruled out upon closer inspection; see below for explanations. As we learn more, we expect to have a growing list of careers that were ruled out.
Developed world mental health professional
80,000 Hours has claimed that becoming a doctor is rarely the best way to do good for someone capable of entering that career, due to the relatively small number of people you can treat directly, and because of the high number of people who will enter this career regardless of whether you do or not.
We think the same is true of people capable of entering mental health professions like psychotherapy and social work. Entering any of these (or similar) paths, a person could treat a relatively small number of people directly through their services. In addition, these are widespread professions that many people would enter anyway. Thus, we think a person could help many more people (and possibly by a greater amount), by instead focusing on options in less crowded areas of the workforce (such as the ones listed in the careers under consideration). For example, it seems plausible that a compassionate individual with appropriate social skills who’d be fit for social work, might also do well in developing world organizations focused on mental health or pain relief. Here, they might be able to put that same set of skills to use, by being involved in the delivery of treatment “on the ground”. In addition, much of this work would likely take place in impoverished areas where their skills would presumably be in (even) greater demand.
3. Methods selection and the career evaluation procedure
Here, we present the procedure for the evaluation of careers aimed at optimising happiness. First, the steps of the process will be listed, whereafter we’ll briefly elaborate on each in turn, in order to make the process more transparent.
3.1 Outlining the career evaluation approach
In brief, the steps for evaluating a potentially promising career will probably include the steps below, which represent a shortened and adapted version of the 80,000 Hours approach to generating high impact options. More specifically, it’s been adapted by our inclusion of WALY’s (the term will be defined below) as a measure of expected value and shortened by leaving out the parts that address how an individual can generate options for themselves (as our first goal, as stated above, is to produce generic and shallow career reviews).
The steps in the (shallow) evaluation of welfare-maximising careers:
1. Crude intuitive judgments about the expected value (in WALY’s), on the margin, of adding one person working in the career
2. Research of the relevant cause area, learning about the following:
a) Identifying bottlenecks to impact in the relevant cause area
b) Learning more about what job roles (including skills) would do the most to alleviate those bottlenecks
3. Reconsidering how promising the career seems, in light of what was discovered.
a) What do we now think the expected value (in WALY’s) would be, on the margin, of adding one person to working in that career?
3.2 Detailing the career evaluation approach
In this section, we’ll explain each of the steps in turn.
Given our commitment to discovering the best career opportunities for maximising happiness, we’ll start the evaluation of a career option by considering its expected impact on well-being adjusted life years (WALYs), for which we use life satisfaction scores: life satisfaction is typically found by asking people “Overall, how satisfied are you with your life, nowadays?” (0 - 10). Measuring WALYs is analogous to the measurement of QALYs and DALYs , the latter two being used to quantify disease burden in the public health domain. We use WALYs as that better captures what we want to improve - happiness. For more about WALYs and measuring life satisfaction and happiness, check out the Happiness Manifesto.
As we don’t expect there to be direct evidence available of what the WALY-value of different careers is, we expect to rely on intuitive judgments - based on our understanding of the happiness literature - to assess if an option seems like it might be high impact. If it does, we’ll flag it as worth investigating further (as was done with the list of “careers currently under consideration” above).
To assess the potential impact on WALYs of a career, we will look at the direct impact we expect someone would have, that is, the impact they would have if no one else would have taken the job.
In many cases, however, we should expect that someone else would have taken their place in that career, and the greater the likelihood of that happening, the lower the impact of entering the career. This is the central intuition behind ‘counterfactual impact’: we want to identify opportunities for making a difference that result in a different outcome than the one we otherwise would have seen.
To incorporate this in our assessment of overall career impact, we’ll adjust our initial assessment of direct impact by subsequently adjusting for counterfactual impact. If it seems likely that the person considering the career wouldn’t be largely or entirely replaced we’ll continue our investigation of that option.
Then, we’ll start researching the relevant cause area (i.e. the problem targeted by the career under consideration). We’ll focus on first identifying the main bottlenecks that seem to prevent necessary progress in that area. These might be lack of funding, skills, insight, information or interest in the problem that the career is working to solve.
We’ll then try to learn more about what sorts of jobs and skill-sets would solve those bottlenecks. For instance, if lack of funding is the key bottleneck preventing progress in global mental health, it would mean that the input most needed to make progress in this area, would be to raise more money. If so, we might consider, say, whether creating jobs that involve lobbying for more funding to mental health NGOs operating in developing countries would be the best way of addressing the mental health burden.
Finally, we’ll return to and adjust (if necessary) the intuitive judgment of the expected impact (in WALYs) of adding one person to working in the career that was considered. The research into the career and its targeted problem area should inform this judgment and make it more reliable than when we formed that initial judgment before learning more about it.
In summary, the approach will consist in making a very rough guess about expected career impact, then investigating more to understand that career and refine our guess about expected career impact.
4. Key uncertainties
Here are what we consider to be the key uncertainties (in roughly descending order of importance):
5. What’s next?
The first career we intend to review is that of direct work in developing world mental health organizations that focus on increasing access to psychological treatments, such as cognitive behavioral therapy.
Following that, careers that we are likely to prioritise reviewing include:
 Happiness means a positive balance of enjoyment over suffering; as such, we can increase happiness by reducing suffering. ‘Happiness’ and ‘welfare’ will be used as synonyms.
 In particular, further research on moment by moment assessments of people’s experiences (the experience sampling method) seems promising.
 QALY: quality-adjusted life year; DALY: disability-adjusted life year.
 Cf. the definition of “key bottleneck” used here.