How this calculator works
Income percentiles compare a household with the rest of the country. Because income is strongly skewed, with a long tail of very high earners, the shape is closer to a log-normal curve than to a bell curve: many households sit near the low end, and a small number earn far more. This calculator fits a log-normal distribution to two published summary measures, the 2023 US household median of $80,610 and the mean of $114,500. The fit reproduces those two anchors exactly, so a household at the median lands at the 50th percentile. Two consequences follow. First, the result is a model, not a lookup: the actual distribution has more mass in the middle and a fatter top tail than a log-normal, so the model overstates how far up the distribution a very high income sits and understates the extreme top. Second, it is household income before tax for the whole country, not for your city, age group or education level. A $120,000 income is well above the national household median, but it can be ordinary in a large city with a high cost of living. Percentiles also say nothing about wealth, which is far more unequal than income. Treat the answer as a national benchmark that is useful for context and not as a ranking of how well off you are.
Formula
Let median M = $80,610 and mean = $114,500. The log-normal spread is sigma = square root of 2 x ln(mean / M). Percentile for an income x = the standard normal cumulative probability of ln(x / M) / sigma. Income at percentile p = M x e^(sigma x z), where z is the standard normal quantile of p. Share of households above the income = 1 - percentile.
Worked example
At $80,610 the model places a household at the 50th percentile. The sigma is the square root of 2 x ln(1.4204) = 0.8375. At $120,000 the z score is ln(120000 / 80610) / 0.8375 = 0.475, which corresponds to about the 68th percentile, so roughly 32% of households earn more. At $40,000 the z score is -0.836 and the percentile is about 20%. The income at the 90th percentile is $80,610 x e^(0.8378 x 1.2816) = about $235,900.
Advertisement
Frequently asked questions
Is this an official government percentile?
No. It is a model fitted to the published median and mean. Official percentile thresholds come from analysing the full microdata. The model is close in the middle of the distribution and less reliable at the extremes.
Is the income before or after tax?
Before tax. The Census measure is money income before taxes, including wages, self-employment income, interest, dividends, rental income and transfer payments such as social security.
Why is the mean so much higher than the median?
Because income is skewed. A small number of very high earners pull the average up while most households sit below it. When the mean exceeds the median, the distribution is tilted toward the high end.
Can I compare an individual income?
Only with care. This model is for household income. An individual's income compared with household figures will look lower than it is, because a household often has more than one earner.
Does a high percentile mean a household is wealthy?
Not necessarily. Income is a flow and wealth is a stock. A household on a high income with large debts can have less wealth than one on a modest income that has saved for decades.