GiveDirectly
Unconditional cash transfers to people living in poverty, so they can decide what their households need most.
The recipient of this project's 50% charity pledge.
Explore & support GiveDirectly ↗● 01 / INCOME & LONGEVITY · UNITED STATES
How much longer do the richest people live?
Nearly 15 more years for men.
About 10 more years for women.[1]
That's the gap between the richest and poorest 1% in this US study. The two lines show how expected lifespan changes across the income distribution.
The gap is wider for men. The poorest men have the shortest lives of any group.[1]
US data, not a global estimate · 2001–2014
Estimates for people who reached age 40, adjusted for differences in racial composition.
Two lines. One income divide.
United States · 2001–2014
MenWomen
Expected age at death, among people who reached 40
Poorest 1%Household income percentile →Richest 1%
Gap between selected groups: men 14.6 years · women 10.1 years
Shading: approximate 95% intervals around estimates, not individual lifespans. Download data ↗
The data could not load. Please reload or download the source data below.
02 / THE WHAT-IFA possibility, not a prediction
Could a smaller income gap
mean longer lives?
Being poor is one thing. Living in a place with a wide gap between rich and poor is another. Could that gap itself affect health?
Researchers found a small link between wider income gaps and a higher risk of dying. But the evidence is uncertain: other differences between places could help explain the link.[2]
ONE EXAMPLE, WITH A BIG “IF”
What could a smaller income gap mean?
Imagine the gap between rich and poor were smaller. If the link found in the study were entirely cause and effect, our example calculation would suggest about 2% fewer deaths, if everything else stayed the same.[2]
Explore the assumptions behind this example below.
Go deeper: try the assumptions & see the math +
What is Gini? It is a measure of how unevenly income is shared. On the scale used here, 0 means everyone has the same income; 1 means one person has all of it. Lower means more equal. For example, going from 0.40 to 0.35 is a drop of 0.05—not a 5% pay rise.
The example above assumes a 0.05-point Gini reduction and treats the entire study association as cause and effect. It is hypothetical, not a prediction or a bound on the real effect.
What does this calculator do? It asks how the result changes when you choose a smaller income gap and decide how much of the study's link to treat as cause and effect. Neither choice is a prediction.
TRY A DIFFERENT WHAT-IF
A drop on the 0–1 Gini scale above. This example is not tied to a particular country.
None: other factors explain the whole link. All: treat the whole link as cause and effect. The study cannot choose this for us; this control changes the assumed share of the log association.
FEWER DEATHS IN THIS WHAT-IF
With no cause-and-effect link assumed, the calculation shows no change.
If the whole link were cause and effect
Using lower / higher study estimates
This range tests uncertainty in the study's link. It does not bound the real effect or promise lives saved.
The formula & its limits
The review's pooled relative risk is 1.02 (95% CI 1.00–1.04) per 0.05 Gini increase. We multiply the log association by your assumed causal share.
fraction = 1 − exp(−causal_share × ln(RR) × Gini_reduction / 0.05)We recalculate using RR 1.00 and 1.04 to test sensitivity. This model assumes the same proportional relationship across inequality levels, that the study's link applies to the place being considered, that the change applies across its population, that other conditions stay the same, and that effects have no time lag. These assumptions are unverified. It does not account for every other factor or identify a specific policy's effect.
The calculator starts at zero assumed cause and effect. That is a starting assumption, not a finding that inequality has no effect. No death count is displayed: a suitable population, period, and mortality baseline have not been validated.
03 / THE BIGGER PICTUREBroader deprivation context
Big claims need
clear definitions.
Oxfam's 2022 “Inequality Kills” report combines healthcare, hunger, and gender-based violence estimates under a broader definition of inequality — not the US income gradient shown above.[3]
Their headline: roughly 21,300 deaths a day from healthcare, hunger, and gender-based violence combined (a separate 2030 climate projection is kept out of this sum, since it isn’t a current death count). This is Oxfam’s own advocacy estimate, not a figure we calculate — it mixes different countries, years, and definitions of inequality than the data above, and should not be read as the effect of any specific income gap.
≈21,300/ dayHealthcare 15,342 + hunger 5,773 + gender-based violence 203 (rounded). Oxfam’s 2022 estimate, not independently verified here.[3]
04 / MAKE IT PERSONALUS household income · private to your browser
Where do I fit?
What could I help change?
See which historical income groups are closest to yours, then explore what sharing a portion of your income could support.
Enter your income, then press “Show my insights” or hit Enter.
Your insights
01 / THE GAP AROUND YOUR INCOME
Your closest groups in the historical study
Your income match on each line is marked with a hollow circle.
These are historical group comparisons, not a prediction of your lifespan or an exact current income rank. The study followed people who had reached age 40.[4]
Explore the full chart ↑02 / WHAT YOU COULD HELP CHANGE
What could your giving do?
per year · per month
Illustration using GiveWell's approximate $7 cost for a child's seasonal malaria-prevention course, published September 2025. Actual costs, funding needs, and results vary.[5]
Explore the evidence & giving options ↗What about lives saved? +
These are statistical estimates across programs and populations, not a count of identifiable people or a promise about your donation. The range compares historical program averages, not uncertainty bounds or today's marginal cost. It isn't an offset against deaths attributed to your income.
Calculation: annual giving ÷ $5,500 to annual giving ÷ $3,500. Check current research and funding needs before choosing where to give.[5]
How we find your place in the study +
Why age 40? The study compared household earnings around age 40 and estimated lifespan for people who had reached that age. We use those historical groups as context; we do not use your age or predict your lifespan.
We adjust your input from September 2026 purchasing power to 2012 dollars using US CPI-U (229.594 ÷ 336.940), then find the closest mean household income at age 40 in each sex group's source table. Means are not percentile boundaries, so this gives a historical group match, not your exact percentile. Changes in the income distribution since the study, your age, household composition, and local costs are not accounted for.
Zero and negative incomes are outside this tool's scope; the study excluded zero-income observations. At the extremes we show the nearest available group without extrapolating. Donation illustrations use your entered dollar amount and historical GiveWell benchmarks; we do not treat giving as a measured reduction in US inequality.
05 / FROM UNDERSTANDING TO ACTION
More income.
More room to build a life.
Three organizations tackling income insecurity through direct cash, community investment, and support for families. Explore their work and choose the approach that matters to you.
UpTogether
Direct, unrestricted cash investments in families facing financial hardship, alongside work toward greater economic opportunity.
Explore & support UpTogether ↗LIFT
Cash assistance plus financial, education, and career coaching for parents working toward greater financial stability.
Explore & support LIFT ↗These income-focused programs are separate from the malaria-prevention example above; we don't apply its impact estimates to them. Descriptions are based on each organization's published program information, checked October 20, 2026.
SUPPORT THE PROJECT
A LITTLE SUPPORT. SHARED.
Help keep this project going.
Consider putting a small part of that planned amount toward this project—and helping someone else directly.
I'll donate 50% of every contribution to GiveDirectly. The other half supports this site, its research, and the ongoing project.
Learn about GiveDirectly ↗Based on your yearly giving plan. This is a one-time contribution, not a subscription.
Choose an amount to see exactly how it would be shared.
- To GiveDirectly
- To the site & ongoing project
- Remaining for charities you choose
The contribution comes out of your giving plan. The impact example above uses only the remaining amount; GiveDirectly's share isn't counted as malaria prevention. Any odd cent goes to charity.
Choose an amount above, then confirm it on Stripe’s secure one-time checkout.
Stripe checkout displays “Warp Cycling,” the creator’s payment account.
One-time payment to this project. I'll pass the charitable share on to GiveDirectly and cover processing fees from the project's share.
PROJECT UPDATESGet occasional updates
Get occasional updates
on this project.
New evidence, corrections, and practical ways to help. Nothing else.
THE SOURCE NOTESEvidence reviewed October 20, 2026
Check the work.
Income and life expectancy
Chetty et al. (2016), JAMA. National Table 1, pooled 2001–2014. Race-adjusted estimates conditional on reaching 40. Gap = selected percentile estimate − comparison estimate; point interval ≈ estimate ± 1.96 × standard error. No interval is inferred for the difference.
Read the paper ↗ Data & CC0 license ↗Inequality and mortality
Shimonovich et al. (2024), The Milbank Quarterly. Pooled mortality RR 1.02 per 0.05 Gini; 95% CI 1.00–1.04. Low-certainty mortality evidence; the review does not establish causality. Our scenario is an illustrative sensitivity model.
Read the review ↗Broader deprivation
Oxfam (2022), Inequality Kills. Advocacy estimates cover multiple forms of inequality. The lower sum includes no climate deaths; the upper components overlap and should not be added together.
Read the report ↗ Methodology ↗Your historical income match
Health Inequality Project Table 1: mean household income at age 40, in 2012 dollars. We match the nearest mean separately for men and women after CPI-U adjustment. BLS annual 2012 CPI-U: 229.594; September 2026 CPI-U: 336.940. No current rank or personal mortality prediction is inferred.
Source table & codebook ↗BLS price data ↗Illustrative donation impact
GiveWell, Our Top Charities, updated September 2025; checked October 20, 2026. Approximate seasonal malaria-prevention cost: $7 per child. Estimated average cost per life saved across its four listed programs for funding directed in 2022–2024: $3,500–$5,500. These are historical modeled averages, not a current donation forecast.
Research & benchmark estimates ↗