The Price of Happiness: Money's Exponential Math That Most People Get Wrong

⬅️ Back to Articles

Matthew Killingsworth published a paper titled The Price of Happiness that resolves a persistent confusion in the money-and-happiness literature. The short version: money and happiness follow logarithmic math, and most people (including researchers), draw the wrong conclusions from it. The full paper is at happiness-science.org/price-of-happiness.

  1. Happiness and the logarithm of income are correlated at r = 0.98–0.99 from $10,000/year to $500,000/year and beyond. This covers experienced happiness (real-time mood sampled via phone notifications) and life satisfaction. It is the most systematic relationship I’ve seen in any social science dataset. There’s virtually no variance left to explain at the group level after log(income).

  2. This means the marginal value of a dollar declines exponentially. An extra $10,000 matters far more to someone earning $25,000 than someone earning $200,000. Plot happiness against raw dollars and you get the familiar concave curve that looks like it plateaus. But a logarithmic relationship never plateaus. If you plot happiness against log(income), it just keeps climbing at a decelerating rate.

  3. The counterintuitive flip: proportional differences matter equally at every level. A 10% raise is associated with the same happiness difference whether you earn $30,000 or $300,000. This follows directly from log math, because log(1.1 × X) = log(1.1) + log(X). The constant term doesn’t depend on X.

  4. Real-world incomes vary exponentially, not linearly. The difference between the 10th and 20th percentile earner is a few thousand dollars. The difference between the 80th and 90th percentile is tens or hundreds of thousands. This means each step up the actual income ladder delivers roughly the same happiness increment. Killingsworth shows income quantile correlates with happiness at r = 0.96, so no log transform needed.

  5. The tension appears when you switch from individual to collective perspective. For your own career, each step up pays exponentially more dollars but delivers constant happiness gains. This is a reasonable trade-off. But for philanthropy, compensation policy, or tax decisions, the exponential math bites hard. A dollar taken from someone earning $73,000 and given to someone earning $2/day generates an estimated 10,000% ROI for collective happiness. A billionaire giving 10% of their income to double the incomes of people earning $2/day would generate roughly 90,000,000% return in happiness terms.

  6. This geometry may explain why income inequality persists. If climbing the ladder feels worthwhile at every rung (constant happiness increment per step), people keep climbing. But since each rung costs exponentially more dollars, the successful capture of those dollars by a few concentrates wealth. Individual rationality (keep climbing) and collective optimality (spread the gains) diverge.

  7. The same math explains why happiness in the US has stagnated despite GDP growth. Income growth has concentrated at the top, where additional dollars have the smallest per-dollar impact on happiness. The people who would benefit most from extra income (the bottom of the distribution), have seen the slowest real growth. If growth is going to be unequal, the optimal distribution for collective happiness is the opposite of what’s happened.

The takeaway: When thinking about your own income, think proportionally, because a 10% raise matters about the same no matter where you are. When thinking about other people’s money (philanthropy, policy, compensation), think exponentially. Giving the same dollar generates radically different happiness depending on who receives it.

Related TMFNK Content

Crepi il lupo!