CEO-to-Worker Pay Ratios: Amazon and Costco Compared

A CEO-to-worker pay ratio of 300 to 1 means the reported executive compensation is 300 times the reported compensation of the median employee under that disclosure’s methodology. It does not mean the chief executive receives exactly 300 times every worker’s salary in cash.

The gap can still be ethically significant. The way to assess it is to read the compensation figures and workforce definition, rather than treating the headline ratio as a complete account of fairness.

Start with the median

Imagine five employees earn 20,000, 25,000, 30,000, 40,000 and 85,000 in the same currency. The median is 30,000: the middle value. The average is 40,000. A fictional CEO package of 9 million would be 300 times the median or 225 times the average.

Those ratios describe the same invented workforce. Switching between them without explanation changes the story. Neither tells you whether the lowest-paid worker can afford local living costs.

What the US disclosure requires

The SEC’s pay-ratio calculation guidance explains the comparison between principal executive officer compensation and median employee compensation, including permitted estimation and sampling approaches. Companies’ workforce geography, occupations and proportions of full-time and part-time staff can affect comparisons.

Use the applicable proxy statement’s own methodology. Read the compensation notes rather than assuming a reported executive package is entirely cash paid during that year. Also check for an unusual award or leadership change before interpreting a sharp annual movement as a permanent change in policy.

Amazon and Costco: a real 2025 comparison

These are each company’s reported 2025 annual compensation figures, in US dollars, from Amazon’s 2026 proxy and Costco’s December 2025 proxy. Their reporting periods and workforce methods are not identical.

Amazon’s population includes full- and part-time permanent and temporary employees worldwide, excluding its CEO. Costco’s main calculation includes part-time, seasonal and temporary employees, with specified country exclusions; approximately 37% of its considered population was not full-time. Its supplemental row changes the employee population, not the CEO’s compensation.

Costco’s main median is higher, yet its ratio is also higher. The ratio and the pay level therefore answer different questions. You cannot infer that Amazon pays a comparable worker better solely because its reported ratio is lower.

Compare the workforce as well as the ratio

A retailer with many part-time roles and an engineering business with highly paid specialists may have different ratios partly because their workforces differ. That does not make either ratio meaningless. It means you should compare like roles and local conditions as well as the aggregate number.

Outsourcing raises another question: are the people doing important low-paid work employees included in the reported population, or workers employed elsewhere? A narrower payroll boundary can make a company appear different without changing the experience of everyone contributing to its operations.

Three measures to put beside the ratio

A low ratio can coexist with poor wages if everyone is paid badly. A high ratio can coexist with decent wages while still raising questions about executive rewards and influence. Avoid using one metric to erase the other concerns.

Turn the number into a better question

Read Amazon’s company research and Costco’s company research alongside the filings. Our figures above are dated 2025 disclosures, not live pay estimates. An annual compensation package also differs from a founder’s total wealth and from cash received when older share awards vest.

The useful question is: “Does the reward structure match the contribution, responsibilities and treatment of people across this business?” The ratio helps you begin that investigation. It cannot complete it.

Research checked 24 September 2026. Examples labelled hypothetical are calculations, not observed company results.