Happiness Statistics Need The Median

I am taking students to Columbia next year and so I saw that it was rated the happiest country in the world according to Gallup polls a couple years ago and is #2 in the current ranking.  This is surprising given the seemingly huge social problems like the long-festering civil war funded by drug cartels in Columbia, but many Latin American countries rate themselves as very happy despite having more obvious social problems than the richer countries of the OECD.  In addition to lower incomes, they suffer lower longevity, more corruption, less educational attainment and higher economic inequality.

Oddly Gallup uses the mean happiness score rather than the median happiness score to measure how happy a group of people are.  This is bad methodology since happiness is not cardinally measurable.  They should use median happiness.

The UN’s World Happiness Report ranks Columbia and other Latin American nations as less happy than the rich OECD nations because it uses a different Gallup poll of happiness data: evaluation vs experience.  Latin Americans tend to rate their life experience higher than people in rich countries whereas richer people tend to evaluate their lives as better overall than poorer people.

The five positive experiences include feeling well-rested, laughing and smiling, enjoyment, feeling respected and learning or doing something interesting; the five negative experiences include stress, sadness, physical pain, worry and anger. The items are grouped into index scores known as the Positive Experience Index and the Negative Experience Index.

People in rich countries don’t feel like they rest, laugh, smile, enjoy life and do interesting things as much as lower income Latin Americans, but rich countries do rate their lives as better.  The life experience measure is a bit more like first-person perspective whereas life evaluation is a bit like looking at life from an out-of-body, third-person perspective.  The World Happiness report prefers ‘life experience’ as a measure of happiness because it is highly correlated with mean income, but that actually makes me more suspicious of the measure.  Mean income should not have as much effect on life evaluation as median income.  For example, the 2013 World Happiness Report notes that the Easterlin paradox is the fact that mean happiness has not risen along with mean US income even though richer Americans report that they are happier than poorer Americans in every given year.

The World Happiness Report suggests that the Easterlin paradox could be explained by the fact that income is not measured property, but they miss the big reason why income has been mismeasured in the US since the 1970s: rising inequality.  We should use median income rather than mean income.  Since 1975, real per-capita mean income has doubled in the US but real median income has been stagnant. No wonder the happiness measures have not risen with mean income. Most people are not earning higher incomes. A small fraction of the population has been getting richer and pulling up the average incomes but most of us are not earning more and so why would the happiness of the masses be correlated with the rise in incomes of the elites?

Unfortunately, nobody has bothered to see how happiness is related to median individual income, but that would be much more important as a measure of group wellbeing than mean income.

Posted in Medianism
2 comments on “Happiness Statistics Need The Median
  1. Vincent says:


    just read your post here. I just had the same thought. It is a joke to measure the mean happiness of countries, so as to measure the mean income.

    Why don’t you write a paper about this, or do you know papers telling about this?

    • Sorry for the late reply, but I’ve been away for a while. I’m working on a paper about median lifetime income as a replacement for mean income if that is what you mean. I don’t have a lot to say about the happiness research because I think it is very difficult to measure happiness so I’m not willing to invest a lot of effort into it, but it is an interesting body of research and what sort of macroeconomic statistic could be more important if we could really figure it out. I’m not optimistic that it is possible to really measure the happiness of a population, but I think the literature occasionally yields an interesting insight.

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 96 other subscribers
Blog Archive