Data currently produced by National Statistical Institutes (NSIs) are often subject to a recurrent revision process, resulting in the publication of successive releases, which represent different quantitative evaluations of the same phenomenon under study. There are various reasons why the revision process is necessary.
For example, revisions can be the result of improvements in the estimation techniques used by NSIs, enlargement of the information set, refinements in the system of basic statistics, changes in the accounting definitions and classifications, or changes in the base year for constant prices evaluations.