Pay by Role editorial · Updated August 25, 2026 · 12 min read
Understanding pay index and cost-of-living index
City pages show a pay index and a cost-of-living (COL) index. Both are scalars you can glance at. Mixing them up is how people “localise” an offer incorrectly or treat COL-adjusted pay as cash an employer will send.
This guide defines both indexes as we use them, including modeled city estimates. For the broader comparison workflow, see salary vs cost of living.
What is the difference between pay index and COL index?
Pay index is about wages relative to a national baseline. COL index is about prices relative to a US-centred baseline of 100.
Two indexes, two questions
| Index | Question it answers | Typical use on this site |
|---|---|---|
| payIndex | How do local wages tend to sit vs national? | Scale a national median when metro percentiles are missing |
| colIndex | How expensive is a typical basket vs a US baseline of 100? | COL-adjusted salary and illustrative rent sketches |
A city can score high on both (expensive and high-paying), high on COL only (pricey without matching wages), or high on pay only (strong wages with moderate prices). The interesting decisions live in those mismatches.
What does payIndex mean here?
For modeled city estimates, we multiply a real national median by the city’s payIndex to approximate local pay when metro-specific percentiles are unavailable. A payIndex of 1.0 tracks the national median; 1.15 implies roughly 15% above national for that modeling path; 0.9 implies about 10% below.
On observed city pages, percentiles come from metro data directly. We may still show how the city median compares to the national median (as a percentage), but that comparison is descriptive — the percentiles themselves are not invented from the index.
Pay indexes compress a messy reality: industry mix, union density, and commuting sheds differ by metro. Use payIndex as a structural adjustment, not as proof that every occupation in the city sits exactly X% above national. Labeling of observed vs modeled is in how we source salary data.
What does colIndex mean here?
colIndex is scaled so that 100 ≈ a US national-average price baseline. A city at 130 is modeled as about 30% more expensive on that basket; a city at 80 as about 20% cheaper.
We use colIndex to compute cost-of-living-adjusted salary: roughly local median × (100 / colIndex). If adjusted pay is higher than nominal, your wages go further against the baseline basket; if lower, prices are eating more of the paycheck.
We also scale illustrative rent sketches from a US baseline using the same index. That is a comparative model for education and exploration — not a scraped listing from a specific neighbourhood.
How should you not misuse the indexes?
- Do not multiply an offer by payIndex to “localise” it yourself when you already have an observed city percentile set — use the observed figures.
- Do not treat COL-adjusted salary as the cash an employer will pay; employers pay nominal currency, not index-adjusted units.
- Do not assume colIndex captures your personal basket. Homeowners, family housing, and international schooling can diverge sharply. Swap in your rent; keep the salary data.
- Do not compare payIndex across countries as if national baselines were identical. Each country’s payIndex is relative to that country’s national wage level. Cross-border comparisons start from local-currency medians (or a common display currency), then apply COL and tax.
What is a simple reading order on a city page?
- Check whether salary figures are observed or modeled.
- Read the local median and percentile band for your role.
- Glance at vs-national percentage and payIndex for wage context.
- Read colIndex and COL-adjusted pay for purchasing-power context.
- Finish with take-home if taxes differ across options you are weighing.
If payIndex is high but COL-adjusted pay looks weak, you are looking at a high-wage, high-price market — negotiate on nominal cash and housing plan. If payIndex is modest but COL-adjusted pay looks strong, the city may stretch a mid-market salary further than a famous coastal metro would.
Key takeaways
- payIndex ≈ local wages vs national; colIndex ≈ prices vs a US baseline of 100.
- Modeled city pay uses payIndex; observed city percentiles do not invent pay from it.
- COL-adjusted salary is a comparison lens, not the employer’s transfer amount.
- Do not compare payIndex across countries as if baselines matched.
- Read observed/modeled first, then median, then indexes, then tax.
Frequently asked questions
- What is a pay index on a salary site?
- Here, payIndex is a scalar for how a city’s wages tend to sit relative to that country’s national baseline. We use it to model local pay when metro percentiles are missing. It is not a published salary, and it is not comparable across countries without converting the underlying medians first.
- What is a cost-of-living index of 100?
- On Pay by Role, 100 ≈ a US-centred average price baseline for the basket we model. Numbers above 100 mean a more expensive modeled basket; below 100 means cheaper. It is not your personal spending, and it is not a consumer-price index from a national statistics office.
- How is cost-of-living-adjusted salary calculated?
- Approximately local median × (100 / colIndex). The result is in the same currency units as the median but rescaled for the index. Use it to rank cities, not to predict next month’s bank balance. Employers still pay the nominal amount.
- Should I multiply my offer by the pay index?
- No, not if an observed city percentile set already exists — use those figures. Even for modeled pages, the site has already applied payIndex to the national median. Multiplying again double-counts. If you are trying to localise a national offer, start from the city page instead.
- Why can a city have high pay and still feel expensive?
- High payIndex and high colIndex together describe a high-wage, high-price market. COL-adjusted pay can look weak even when nominal medians look strong. That is a housing and prices problem, not a bug in the wage data. Negotiate cash and plan rent; do not assume the index “cancels out.”
- Are these official World Bank or BLS indexes?
- No. They are site-level scalars we use for modeling and comparison. World Bank series appear as country economics (for example CPI). BLS-style occupation wages feed salary percentiles through our aggregators. Do not cite payIndex or colIndex as official statistics.
Next step
Indexes are shortcuts. Read them after you know whether the salary is observed or modeled, and never confuse a COL-adjusted number with the cash on an offer letter.
Open a city page and read both indexes →
Then run the same salary through the cost-of-living calculator.