I propose a simple and formal method to reconstruct falsified election results, based on the hypothesis of the mutual independence of the overall turnout and the result of power. This method is a development of the classical method of S.A. Shpilkin, however, it does not require additional assumptions and has no free parameters. In addition, it uniformly reconstructs all key values and uses a simpler and clearer mathematical procedure. Comparison of official and true results of federal elections allows to perform a political science analysis of the development of the institution of electoral frauds in Russia in the 20th century and its impact on the political system.

elections, frauds, electoral statistics, big data, reconstruction, inverse problems, regression analysis

Podlazov Andrey Viktorovich,  ,  orcid.org/0000-0002-0328-3814,  KIAM RAS