Random Tabular Adjustment is here!

March 26, 2019

Balancing confidentiality protection and user needs

Statistics Canada has been exploring new methods to balance the need for more high-quality data outputs with the protection of confidential information in the release of economic data estimates.

Random Tabular Adjustment (RTA) is a statistical technique used to protect confidential information in published data by applying random adjustments to sensitive estimates. This is a new method developed by Statistics Canada and we are applying it for the first time with the release of the Survey of Innovation and Business Strategy 2017. As a national statistical agency, we continuously look for ways to improve and modernize, using new technologies to add value to the data we publish in order to better serve Canadians.

Making noise: Releasing more data with RTA

This new statistical method will allow Statistics Canada to release more useful statistical information, without making any compromise on the protection of confidentiality. Our users have been requesting more detailed information and we expect that they will welcome this update to the methodology.

The RTA method adds random "noise" to cells, which results in preventing the disclosure of individual estimates. This method differs from the suppression techniques that have traditionally been used with economic data tables released by Statistics Canada because it improves the utility of the tables. With RTA, cell estimates are releasable while still ensuring individual values are not disclosed. 

So using this method allows us to increase the amount of useful data we can publish, while at the same time ensuring confidential information is protected. This makes for a great balance.

While the Survey of Innovation and Business Strategy 2017 release is the first to use RTA, the plan is to determine what other datasets could benefit from this new method and to apply it to those in the future.

RTA and economic data

The reason RTA allows for more data to be released is precisely that: instead of "suppressing" the data, the estimates are "perturbed." This means that the sensitive information is randomly altered to protect the confidentiality of the individual responses contributing to the estimate. This is done by adjusting the estimate in question so that a precise value cannot be assigned to an individual contribution. The added benefit of this method is that it does not affect other cells, which are not considered sensitive cells. Only the cells with the sensitivity issue and their aggregates are affected. For more technical information on RTA, please see the paper by Stinner (2017) on the Statistical Society of Canada's website.

Statistics Canada's mandate is to publish data on the economic, social, and general conditions of the country and its people.  In order to do this, it must collect information from individuals and businesses, with the time honoured commitment to confidentiality as enshrined in the Statistics Act and the Privacy Act. Statistics Canada is committed to continuous improvement and looking for ways to provide more data to our users while at the same time protecting privacy and confidentiality.

Login/register to post comments.

Please note that comments are moderated. It may take some time for your comments to appear online. For more information, consult our rules of engagement.

Date modified: