Can we measure happiness in a survey?
The difficulties of measuring self-reported happiness

9 minutes read - 1780 words

Happiness researchers are often required to determine the level of happiness of a population in order to evaluate policies that might affect it. One way to determine population happiness is through surveys. In this post, we discuss some points that one must keep in mind when designing a happiness survey.

If you like reading about philosophy, here's a free, weekly newsletter with articles just like this one: Send it to me!

How do you measure happiness?

We all want to be happy. Our governments often will take decisions in the name of our happiness or welfare. But in order to know whether one makes the right choice, the choice that improves one’s happiness, one must be able to compare two states of affairs and reliably decide which of the two is “happier.” That is, one needs to be able to measure happiness in some reliable way. How can this be done? How can happiness be measured?

A survey

Let’s begin with a fun exercise: Assume that you wanted to measure someone’s degree of happiness. Go on and prepare a questionnaire with 1-5 questions with which you would measure the level of happiness of some random person on the street!

Read on after you’re done.

Let’s see. Look at your survey now, and try to answer the following questions:

Did you think of particular measurements more than once? Are some of your survey questions similar to others? What were the most common questions you thought of?
What understanding of happiness is reflected in your questions?
How could the answer lead to a wrong result, for example by measuring something other than “happiness”?
Is the understanding of happiness reflected in the questions complete, or does it only address a specific version or subset of what we generally call “happiness”?

Here are some questions my students regularly come up with. Try to see how we can evaluate them using the four criteria above:

How often do you laugh loudly? (each month or year)
Do you have any religious background?
Are you a confident person?
Do you live under high pressure?
Are you satisfied with your life now?
How often do you smile or laugh?
Do you have good relationships with your family and friends?

What can we say about each of these questions?

Laughing

Number (1) seems to assume that laughing is correlated to happiness. On a superficial level, we certainly do expect laughing people to be happier than crying ones. But can we really say that people who don’t laugh are unhappy or less happy? One may be perfectly calm and happy in a satisfied way, without ever laughing out loud. While someone may make a big show of laughing his heart out, but in reality they may be unhappy, depressed, and feeling lonely.

Religiosity

Question (2) is also based on an interesting assumption, namely that there is some correlation between religiosity and happiness. We have discussed elsewhere that this is indeed the case. But why would this be so? A few quick thoughts:

Religious people, particularly Christians or Muslims, have the hope of a better life after death, which might make it easier for them to bear hardships in this life.
Christianity emphasises forgiveness, tolerance, and kindness as virtues of the individual (never mind that church history is a stark demonstration of quite the opposite attitudes). These virtues also tend to make their carrier himself happier than, say, greed, vengefulness or anger.
Religions are generally practiced in communities, the members of which display these virtues towards each other. Particularly during hard times, and in poorer environments, having good social relations inside one’s religious congregation or reference group can be of significant practical help to the individual. If you are interested, read about these things in more detail here.

Religion and Happiness

Religion has a profound effect on happiness. Multiple studies have shown that religious believers are generally happier people, an effect that is more pronounced in poorer countries.

Confidence

Question (3) assumes that confident people are more likely to be happy than insecure ones. Is this the case? We can probably imagine counter-examples: a confident sales-man might not be really happy, while someone who is more of a quiet and introvert person might be perfectly happy with his life. But the opposite also has some plausibility: confident people are perhaps less likely to be lonely, they will likely have more friends and perhaps more social recognition, and these factors also tend to contribute to happiness.

Pressure

Question (4) asks whether one lives under “high pressure.” If you think about it, it’s not entirely clear how one would know whether this is the case. Of course, one feels the “pressure” that comes from one’s job, or the environment’s expectations. But consider the “pressure” felt by a Nobel-prize nominated researcher, who now has to make sure that his research results look good; and, on the other hand, the “pressure” felt by a poor man who is trying to earn enough money day-to-day in order to feed his family of seven. These are very different kinds of “pressure,” with different consequences attached to them. Is not getting a Nobel prize equally stressful as being responsible for one’s starving children? Will failure in both cases affect one’s happiness in the same way? Can the “pressures” caused by these situations even be compared to each other in any meaningful way?

Life satisfaction

Question (5), whether one is satisfied with one’s life right now, changes the game a bit. Instead of asking about objective features of one’s life, this is a self-evaluation question. We talked about self-evaluation previously too. Essentially, self-evaluation has the advantage that it allows people to bring in their own criteria for happiness. One person might be happy when playing chess, while another might be happy hiking, or even when suffering for a political cause. Self-evaluation allows all these different conditions for happiness to be recognised as ways that can make one genuinely and personally happy. In contrast, imposing a researcher’s view on what are supposed to be “objective” conditions that make people happy is much less flexible, and may lead to misunderstandings: imagine, for instance, a researcher assuming that happiness is correlated with high income. What now if this researcher is interviewing a monk? Monks don’t have personal income, but are they therefore sure to be miserable? Obviously not.

Note also that question (5) asks about one’s happiness “right now.” This is interesting because happiness changes over time, so it seems good to ask about happiness at a particular moment. But, of course, this can also go wrong: what if I was happy for weeks and weeks until just yesterday, when I had a car crash? Now I’m unhappy, but clearly this state is not representative of my general level of happiness. Also, we know that the weather, for example, can influence questions of this type. Respondents generally report higher levels of happiness on sunny days, or when they accidentally find a small coin on the street (Kahneman 2006). This means that little chance occurrences like that can already change the survey result: can we then still say that the survey is dependable?

Smile!

Question (6) introduces another interesting concept: the fake smile. Obviously, some people have been smiling all day yesterday, and the day before, and all the working days of the past year: cashiers in super-markets, for example, or any other salesperson. Does this mean that they were happier than, say, office workers who weren’t required to smile all day? – Clearly not. We wouldn’t count a forced smile towards one’s happiness! So the question is, can we distinguish a fake smile, one that does not express happiness, from a genuine smile, one that does?

Actually, we can. French neurologist Guillaume Duchenne (1806-1875) discovered that there are two kinds of smiles, depending on which facial muscles are used to produce them. The fake, sometimes called “Pan-Am” smile, is a smile which involves only the muscles around the mouth: we all know it well from salespeople, politicians, and partners in the terminal stages of relationships. The genuine smile, the so-called Duchenne smile, involves also the muscles around the eyes. These muscles cannot be contracted voluntarily; a smile that employs them correlates well with genuine happiness. So if question (6) was rephrased to ask specifically about the Duchenne smile, then it would be a good measure of (at least) momentary instances of happiness throughout the day. It still wouldn’t help us measure the more long-term aspects of life satisfaction that are not usually expressed in smiles.

Family and friends

Question (7) asks about one’s relationship to family and friends. This is indeed another good question. We know that good relationships to family and friends correlate strongly with happiness and life satisfaction (e.g. Easterlin 1974).

Unhelpful questions

And here are some questions that are not that good for a survey like the one we are talking about. Have a look and try to see if you can spot the problem:

(A) Do you think smiling can make people feel happy?

(B) Would you recover soon after a bad event?

Question (A) does not actually evaluate the respondent’s own happiness at all. It asks about the respondent’s _opinion, _ which is an entirely different matter. I can think that smiling makes people happy, and at the same time feel miserable myself.

Question (B) does ask about the respondent’s feelings, but the feelings in question are not correlated to the respondent’s happiness in any obvious way. Is recovering soon after bad events a sign of happiness? Or is it a sign of indifference towards others and towards the world, and therefore a sign of deep-seated unhappiness and emotional dysfunction? It is impossible to say. So when we are asking in surveys about objective facts in order to evaluate happiness, we need to make sure that these objective facts do have some known and strong correlation to happiness; otherwise we are collecting possibly irrelevant or misleading facts.

That’s it for the moment. We had a look at various ways one could come up with in order to measure people’s happiness in a survey. In the next post we will talk about how professionals go about measuring happiness in (more or less) scientific ways.

_This is part 2 of a series of posts on happiness. Find all the posts on happiness here.