The Administrative Council for Economic Defense (CADE) recently dismissed an investigation regarding the provision of benchmarking services involving salaries.
CADE’s initial concern stemmed from advertising campaigns by companies offering access to databases of employee compensation and benefits across different employers. According to CADE, sharing sensitive information about employment terms could lead to salary uniformity.
To assess potential violations, CADE contacted the National Data Protection Authority (ANPD) and the benchmarking service providers. The investigation revealed diverse data sources used by these companies, including third-party data, public and private data, and that they have different business models. CADE also concluded that there were no competition concerns: the information provided by the companies contacted did not allow a market agent to extract specific and sensitive information about their competitors – as the data was presented in an aggregated and anonymized manner.
In this sense, salary data was presented in non-specific formats, such as:
- Quartiles: based on a reference value, a percentage is indicated as being above or below the reference value. E.g., the 1st quartile indicates a value where 75% of the reported salaries are higher than this level and 25% are lower;
- Averages: the value of the sum all reported salaries for a position divided by the number of reports; or
- Ranges: presenting the lowest and highest reported value.
Regarding companies, the data could be aggregated considering:
- Size or Revenue: E.g., managers in small companies or up to a certain revenue receive salaries within a specific range;
- Region: E.g., managers in São Paulo receive salaries within a certain range.
These aggregation methods prevented identifying individual salaries and/or specific companies.
Therefore, CADE’s recent decision solidifies key guidelines for sharing compensation information, such as: i) not individualizing salary values – aggregating this information in ranges, averages, or quartiles; and/or ii) anonymizing the companies to which the remuneration refers – aggregating them by size or region. These precautions ensure the legality of data sharing.
