What Makes People Happy? What Fifty Years of Research Actually Found

Researchers separate two things that the word happiness runs together. One is how a person feels from moment to moment, the balance of pleasant and unpleasant emotion across a day, which is measured by asking people to report their mood at random times or to reconstruct yesterday hour by hour. The other is how they judge their life as a whole, the evaluation that a single question about satisfaction captures. The two are related and come apart in instructive ways: parents of young children report more stress and fatigue moment to moment and higher life satisfaction overall, and a demanding job can score badly on the first and well on the second. Most of what follows concerns the second, since that is what the large surveys measure, and it is remarkably stable, reliable and comparable across countries.

Money matters, and less than people think. Within a country, richer people are on average more satisfied, and the relationship is logarithmic: each doubling of income adds about the same amount of satisfaction, so the gain from 20,000 to 40,000 is like the gain from 100,000 to 200,000, which means that the poor gain far more from an extra pound than the rich. Whether the effect flattens entirely at some level has been argued since a 2010 paper put a ceiling on day-to-day mood at about 75,000 dollars; later and larger data suggest that satisfaction keeps rising slowly, that mood flattens for most but not all, and that for the unhappiest fifth of people money stops helping once basic needs are met, because what ails them is not financial. Between countries, richer ones are happier, and the countries at the top of the rankings, the Nordic states, Switzerland and the Netherlands, combine high income with trust, low corruption, and welfare states that limit the fall.

Across surveys, longitudinal studies that follow people for decades and the natural experiments of unemployment, marriage, bereavement and lottery wins, a few factors dominate:

• •Relationships: having close, trusted people is the strongest single predictor, and the Harvard study that has followed men since 1938 concluded that the quality of relationships at fifty predicted health and happiness at eighty better than cholesterol did
• •Health: mental health above all; chronic pain and depression are the heaviest weights, and physical illness matters most when it limits what a person can do
• •Work: unemployment is one of the largest and most lasting blows, larger than the income loss explains, and job security, autonomy and a sense of purpose count more than pay
• •Freedom and trust: living in a society where one can choose how to live and expect a lost wallet to be returned
• •Genes: twin studies put about a third to a half of the variation between people in temperament, which sets a baseline each person returns to

People adapt to most changes faster and more completely than they predict. The paraplegic and the lottery winner in a famous 1978 study were, after a year, closer in happiness to each other and to everyone else than anyone expected; a new car, a bigger house and a promotion fade within months; and the same is true, mercifully, of most losses. Adaptation is incomplete for some things, unemployment, chronic pain, loneliness and the loss of a spouse among them, and it is fastest for things one stops noticing, which is why spending on experiences, which are remembered and shared, tends to leave more behind than spending on objects, which are got used to. The failure to anticipate adaptation, called affective forecasting error, is the main reason people pursue what will not make them happy and fear what they would survive.

The interventions with replicated effects are modest and unglamorous: regular exercise, which in trials rivals antidepressants for mild depression; enough sleep; time with people one likes, deliberately protected; practising gratitude, in the specific form of writing down a few good things each week; acts of kindness, which raise the giver's mood more reliably than the receiver's; and time spent in nature. The effects are small individually and add up. What does not work, or works less than expected, is the pursuit of happiness as a goal in itself: people who rate being happy as very important to them tend to be less happy, apparently because they monitor their mood and find it wanting, and the more durable satisfactions turn out to be by-products of doing something else, work that absorbs, people who matter, and a purpose that outlasts the day.

Research that asks people how satisfied they are finds that income raises happiness logarithmically and less than expected, that close relationships, mental health, meaningful and secure work, freedom and trust predict it most strongly, and that genes set a baseline to which people return after most changes because adaptation is fast for possessions and slow for loneliness, unemployment and pain. The practices with real effects are small ones, exercise, sleep, gratitude, kindness and protected time with others, and happiness pursued directly tends to recede.