How Accurate Are Open-Source Social Security Calculators?

People use open-source calculators to estimate their Social Security retirement benefit, and no study we know of checks them against an exact answer. We built the Social Security Administration's published calculator, ran 1,231 synthetic (made-up) workers through it, many on the edges of the rules, and used its results as the answer key. A second program, written from the law, agrees on every worker. Of five calculators, one matches the agency for every worker. Three give a wrong monthly benefit for 2, 15 and 19 percent of workers, mostly through three small mistakes. Two round to the nearest cent before rounding down, making some benefits a dollar too high. Two miss that the law counts someone born on 1 January as born the year before, worth up to 175 dollars a month. One ignores the yearly limit on countable earnings, putting a high earner typically 597 dollars a month too high. An ordinary worker, born mid-month and earning under the limit, gets the right amount or a dollar more. The fifth counts only earnings from age 22 to 61, as its code states, and runs a little low. Why were the mistakes not caught? Developers test against published worked examples, and three, in a Congressional Research Service report and the federal regulations, print an amount the rules do not give. Checks written without an answer key caught the two largest mistakes and missed the rounding one. We release the workers, official answers and checks so anyone can test a calculator this way.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24376857
Primary Topic
E-Government and Public Services
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

How Accurate Are Open-Source Social Security Calculators?

Shirshak Jung Thapa
Zenodo (CERN European Organization for Nuclear Research)
E-Government and Public Services
preprint

How Accurate Are Open-Source Social Security Calculators?

Shirshak Jung Thapa
preprint en

Abstract

People use open-source calculators to estimate their Social Security retirement benefit, and no study we know of checks them against an exact answer. We built the Social Security Administration's published calculator, ran 1,231 synthetic (made-up) workers through it, many on the edges of the rules, and used its results as the answer key. A second program, written from the law, agrees on every worker. Of five calculators, one matches the agency for every worker. Three give a wrong monthly benefit for 2, 15 and 19 percent of workers, mostly through three small mistakes. Two round to the nearest cent before rounding down, making some benefits a dollar too high. Two miss that the law counts someone born on 1 January as born the year before, worth up to 175 dollars a month. One ignores the yearly limit on countable earnings, putting a high earner typically 597 dollars a month too high. An ordinary worker, born mid-month and earning under the limit, gets the right amount or a dollar more. The fifth counts only earnings from age 22 to 61, as its code states, and runs a little low. Why were the mistakes not caught? Developers test against published worked examples, and three, in a Congressional Research Service report and the federal regulations, print an amount the rules do not give. Checks written without an answer key caught the two largest mistakes and missed the rounding one. We release the workers, official answers and checks so anyone can test a calculator this way.

Zenodo (CERN European Organization for Nuclear Research)
E-Government and Public Services
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon