The importance of data quality
The quality of vital statistics data is closely related to how well the data cover the entire national territory, and whether all vital events taking place are recorded and registered. The aim of any civil registration system should therefore be to have universal coverage – that is, to have functioning registration facilities available to the entire population. To know how well a national system is performing, information is required on both coverage and completeness – that is, the extent to which the population has access to vital registration, and the extent to which vital events are registered or notified in the population as a whole. The extent of coverage and completeness very much determine the overall reliability and utility of the tabulations produced.
Coverage is the proportion of the population that the civil registry system serves. Coverage can be measured by counting the population in the administrative areas that have access to local registration as a fraction of the total population.
Conversely, lack of coverage can be measured by counting the population in those administrative areas that have no local registration facilities available to them as a fraction of the total population of the country.
The terms coverage and completeness are often used inter-changeably and in some languages coverage actually means completeness.
Examples of coverage
Imagine a country where 30 per cent of the population live in urban areas, all of whom have access to registration facilities, but where an estimated 80 per cent of deaths in those urban areas were registered. Suppose also that there are no registration facilities available to the rural population. Then the coverage of death registration in this country would be 30 percent, and the completeness of death registration (in the covered population) would be 80 per cent. Note, however, that the overall completeness of death registration in the country would be much less than 80 per cent, since all the deaths occurring among the rural population would have been missed. So if, for example, the actual number of deaths in urban areas was 50 000, of which 80 per cent (40 000) were registered, and the actual (estimated) number of deaths in rural areas was 150 000, then the completeness of registration at the national level (as opposed to the covered, urban population) would be 40 000/200 000, or 20 per cent.
United Nations (2014). Principles and recommendations for a vital statistics system, revision 3, UN Department of Economic and Social Affairs, New York, http://unstats.un.org/unsd/Demographic/standmeth/principles/M19Rev3en.pdf.