Research units A-Z

Workstrand 6 provides insights into how the general public and policy makers value different policy outcomes.

For example, if you had to choose between a ‘Policy A’ that leads to greater income and a ‘Policy B’ that leads to better health, which would you pick?  Or, between a ‘Policy C’ that leads to greater average population health but with greater health inequality and a ‘Policy D’ that leads to lower average population health but with less health inequality?

These are difficult choices and can invoke emotive reactions. It would be nice if we could increase income, increase health, and reduce inequalities. But in reality, we often cannot do everything that is good, and policy makers have to decide between policy options.

The balance between different outcomes is known as a trade-off – in these examples, better health versus higher income; or improving the average level of health versus reducing inequality in health. Understanding how people think policies should trade off between different outcomes is important for SIPHER because it lets our policy partners reflect public preferences when deciding between policy options

To provide insights to societal valuations Workstrand 6 looks at two things.

• How to best measure a person’s overall wellbeing in a single number; and,
• How to adjust the index of population wellbeing by how much people are concerned about inequality.

1. Measuring overall wellbeing

Policies aim to improve population wellbeing, in terms of better health, more income, better housing, safer neighbourhoods and so on – these outcomes are different contributing factors (called “domains”) that make up overall wellbeing.

In SIPHER, we have identified seven domains of wellbeing that we will focus on.

• How physical health affects daily activities
• How emotional problems affect daily activities
• Feeling lonely and left out from others
• Household spending money
• Employment situation
• Concern about the safety of the neighbourhood
• Quality of the housing

This means that the wellbeing of any person can be described as a combination of these seven domains, which gives us rich information about how people are doing. But it also means sometimes we will not be able to say if a new policy would improve or worsen somebody’s wellbeing. Suppose a policy leads to somebody getting a new job that pays better but is also very stressful. If at least one of the seven domains go up but one or more come down, what is the overall effect? We cannot cancel out one worsened domain with one improved domain, because the changes may not carry the same weight.

This example illustrates the big advantage to having a measure that summarises across these different domains into a single index of wellbeing, by reflecting the relative importance of each domain. A single wellbeing index makes it possible to compare across different policies and rank them according to how good they are at improving overall wellbeing.

To do this, we have conducted a study to examine how people trade off between the seven domains of wellbeing. Members of the general public helping us with this study were asked questions about their personal preferences (that is, talking about themselves) for different combinations of the seven wellbeing domains. We have also asked the general public about their social preferences, or how they think policy makers should weigh up the different wellbeing domains that other people may experience. For example, if policy makers could bring about two different combinations of wellbeing domains, which one should they go for?

Once we know how people trade off between different domains of wellbeing (either personally or socially), we can use these results to build a single wellbeing index that summarises any combination of the seven wellbeing domains in a single number in a way that reflects the relative importance of each domain.

2. Adjusting the index of population wellbeing for how much people are concerned about inequality

As well as concern for different outcomes (e.g. greater income, better health and safer neighbourhoods), the general public and policy makers are interested in how different outcomes are spread across society, and whether people regard that spread as fair. For example, we know that different groups of people enjoy different levels of overall wellbeing. To what degree do we accept such differences, and to what degree do we want people in society to enjoy similar levels of wellbeing? Note that achieving similar levels of wellbeing for everybody might only be possible if we settle for a lower level of average wellbeing.

We are hosting discussion groups to better understand how people think about inequalities. By doing this, we will develop a measure of “inequality aversion” – this captures how much people dislike inequality. It works by measuring what lower level of average wellbeing people would accept, if it was combined with complete equality. Those who strongly dislike inequality would accept a pretty low level of average wellbeing provided complete equality was achieved. On the other hand, those who are more relaxed about inequality would not want to sacrifice much average wellbeing just to achieve complete equality.

Knowing people’s inequality aversion will help us decide between policies that are better for overall wellbeing but make inequalities worse, and policies that are good at tackling inequalities but may result lower average wellbeing. Put another way, a question could be if people prefer a policy that can increase total wellbeing of the population by 10% through focusing improvements on those who are already better off (and thus increase the inequality), or a policy that can increase the total wellbeing of the population by only 1% but focuses these improvements on those who are currently the worst off (and thus reduce the inequality).

The research in Workstrand 6 is enabling SIPHER to help policy makers deal with difficult and potentially emotive issues by offering a transparent and consistent approach to:

• Rank policies that have different impacts on different domains of wellbeing. This is done by calculating which has larger wellbeing improvements: e.g. Policy A that leads to better household finances and employment outcomes but poorer physical and mental health outcomes, or Policy B that does the opposite. And then,

• Adjust the above ranking of policies by considering not just how much overall wellbeing a policy improves, but also what impact it has on inequalities across people. This is done by taking into consideration how strongly people dislike inequality.