Replication: Income and emotional well-being: Evidence for well-being plateauing around $200,000 per year

Replication Analysis

Income and emotional well-being: Evidence for well-being plateauing around $200,000 per year
Mikkel Bennedsen — 33391 observations

Summary

Original finding (KKM2023, threshold at $100,000): Emotional well-being increases monotonically with income without a plateau. Slope above threshold: 0.1088 (SE: 0.0205) — positive and similar to below-threshold slope.

Replication (Bennedsen, data-driven threshold at $175,000): Emotional well-being plateaus above ~$175,000/year. Slope above threshold: -0.0015 (SE: 0.0541) — essentially zero.

Step 1: Finding the Optimal Threshold

The sum of squared residuals (SSR) is computed for each candidate income threshold. The threshold that minimizes SSR gives the best-fitting structural break point.

Step 2: OLS Regression Comparison

Piecewise linear model: E(z_i|x_i) = (a + b·x) below threshold, (c + d·x) above threshold. The z-scored well-being is the dependent variable; log-income is the independent variable.

	$100,000 Threshold		$175,000 Threshold
Regime	Slope	SE	Slope	SE
Below threshold	0.1121	0.0133	0.1108	0.0088
Above threshold	0.1088	0.0205	-0.0015	0.0541
SSR	33137.90		33132.75

$100,000 Threshold

$175,000 Threshold (Data-driven)

Step 3: Quantile Regression Comparison

Quantile regression examines different parts of the well-being distribution. t-statistics above ~2 indicate statistical significance.

$100,000 Threshold

Quantile	Slope below	t	Slope above	t
15%	1.900	8.13	0.335	0.92
30%	1.329	6.86	1.228	4.10
50%	1.237	6.46	1.459	4.94
70%	1.178	5.32	1.907	5.60
85%	0.791	2.96	1.983	4.87

$175,000 Threshold

Quantile	Slope below	t	Slope above	t
15%	1.826	11.67	-0.036	-0.04
30%	1.344	10.45	0.381	0.48
50%	1.233	9.73	0.147	0.19
70%	1.153	7.89	0.547	0.61
85%	1.003	5.62	-0.354	-0.32

Quantile Regression: $100k Threshold

Quantile Regression: $175,000 Threshold

Conclusion

The replication confirms the paper's central finding: the conclusion that emotional well-being increases monotonically with income (KKM2023) is sensitive to the placement of the income threshold. When the threshold is chosen in a data-driven way (minimizing SSR), it shifts from $100,000 to $175,000, and the relationship above that threshold becomes flat — both in the OLS analysis and across all quantiles of the well-being distribution.