Why "3–6 months" is too crude to use
The 3–6 month rule was popularized in the 1970s. Today's labor market is segmented — a software engineer in Seattle has a very different job-search distribution than a hospitality worker in a tourist town. A dual-income household with a 1-month emergency fund is statistically more resilient than a single-income household with 6 months. The math adapts; the rule of thumb doesn't.
FAQ
Why does industry matter so much?
BLS JOLTS data shows that healthcare and government roles have an average duration of unemployment under 20 weeks, while media and hospitality routinely run 30+ weeks. The difference between 5 and 8 months of runway is the difference between 70% confidence and 95% confidence.
Should I include severance in the runway calculation?
Yes, if you have a written severance policy at your employer. Most white-collar roles come with 1–3 months; some come with much more. Subtract the expected severance from the target. This calculator doesn't model severance explicitly because it varies so much — adjust the "monthly expenses" down by the severance amount if applicable.
Where should I keep the emergency fund?
Somewhere you can move it to checking in 1 business day. A high-yield savings account (HYSA) is the standard answer — current US rates are 4–5%. Money-market funds at a broker are similar. Do not put it in stocks or long-term bonds; the whole point is that the value is stable when you need it.
Methodology
Job-search duration is modeled as a lognormal-ish distribution: high probability of short searches, long tail for drawn-out ones. The base median is 3 months. The industry multiplier scales the median and the tail. Household structure and liquidity multipliers further adjust. The target is the duration at the chosen confidence quantile of that distribution, multiplied by monthly essential expenses. The chart shows the CDF flipped into a duration distribution. Base parameters are calibrated to BLS JOLTS / EU-LFS data; the multipliers are derived from observed industry-level median duration of unemployment (2022–2024 vintage).