Ukraine - Subnational Population Statistics

A major limitation of this COD-PS is its reliance on a population census conducted more than 20 years ago, namely the 2001 population and housing census of Ukraine.

The 2001 census in Ukraine was subject to severe undercount in all oblasts — with an average undercount of 10.5%. This undercount was more severe in large urban areas (e.g. Kyiv City, Kharkiv, Sevastopol and Dnipropetrovsk). No sex differential was observed in the 2001 census undercount. Raw census data have been adjusted for undercount in the census, along with updated birth and death registration data, prior to the projection of the population forward from 2001 to 2020. The census undercount was concentrated in the age group 0-4 years and ranged from 7% in Transcarpathia to 15% in Sevastopol and Kharkiv and 16% in Kyiv.

UNFPA has assessed population age-/sex-specific projections against available household survey data (namely the 2014 Household Budget Survey, 2012 MICS Survey, 2007 DHS Survey and 2009-2014 Labor Force Survey). These ADM-1 population projections are consistent with the fertility and mortality trends documented in the available survey data. The vital registration data and ADM-1 population projections for 2020 do not exhibit any notable age heaping.

Birth and death registration in Ukraine is mandatory. The vital registration system in Ukraine between 2001 and 2020 appears to have been well-maintained. The birth and death registration data is approximately complete in Ukraine for this time period in areas that have been under effective control of the government of Ukraine. Since 2014, the completeness and quality of vital event registration in Donetsk and Luhansk has detioriated.

A limitation of the vital registration data in Ukraine is that these data are published by place of registration, not by place of occurrence (and place of registration). As a result, the 2020 population projections using 2001 census data and 2001-2019 vital registration data implicitly assume that there is negligible difference between the place where births and deaths occur and where they are registered. While this is likely true in most urban areas of Ukraine, it is unlikely for the more rural areas of Ukraine where people seek health care (both reproductive care, emergency care and chronic disease care) in urban areas. As a result, this COD-PS is likely to under-estimate the older age population size in rural areas of Ukraine and over estimate the young age population size in urban areas.

This COD-PS assumes that internal migration between oblasts and international migration after the 2001 census are accurately reflected in administrative data on migration flows maintained by the State Statistics Service of Ukraine. Further refinement of the COD-PS is likely needed to account for recent migration movements related to armed conflict. So, where possible, this COD-PS dataset should be integrated with available data on migration stocks and flows from UNHCR, IOM, and other sources.