Population data quality checks: Romanian adult deaths and lives, an evaluation

Abstract

<abstract> <p>This article investigates the quality of the Romanian population data for the years 1966–2018. The following age heaping measures are being used, each with its recommended population age span: Whipple, Myers, Bachi, UNASA, Kannisto, Coale and Li. In addition, the Kannisto measure of old age overstatement is used. Our analysis corroborates the results of our measures with the demographic characteristic of the Romanian population. We aim to establish whether; (a) the data spanning the time interval 1966–1989 is more accurate than the data collected during the 1990–2018 interval, (b) the Romanian data is at least of comparable quality with the tier 2 HMD data and/or (c) the Romanian data is not grossly inaccurate and therefore suitable for research. We find support for (a) and, cautioning against the live data for the time interval 2002–2018, for (b) and (c). Limitations aside (no through overstatement or cohort analysis), this is the first study to attempt such evaluation and hopefully not the last.</p> </abstract>