A/HRC/45/14 responses, including in precarious situations, such as emergency situations, where the needs and rights of older persons are more likely to be overlooked.14 34. Data needs not only to be disaggregated by age, but also by other critical dimensions, including sex, disability, marital status, household or family composition, and type of living quarters, in order to achieve a more granular and meaningful data analysis to inform policies affecting older persons. There is a need to review existing disaggregation protocols for data on social welfare, violence, public participation and other essential indicators to ensure full older-age disaggregation of crucial data, and to promote the generation and tabulation of available data on older persons by 5-year age groups.15 35. In this context, the use of small sample sizes renders disaggregation of data on older persons difficult. General surveys that collect information on older persons often use sample sizes in older age groups that are too small to allow disaggregation by the most basic sociodemographic variables and to expose life-course specificities, such as differences between the third and fourth age. Similarly, longitudinal surveys often fail to include older age groups, leading to significant data gaps on the specific challenges at the intersection of older age with other dimensions such as gender or socioeconomic status. 36. There are other challenges to accurate data collection on older persons that must be considered. Definitions of old age, for example, determine the way in which data is collected, needs are assessed and responses are shaped. For the purposes of law, policy and data collection, old age and older persons are typically determined in merely chronological terms. Bearing in mind that age is a social construct, definitions based only on chronological age fall short of local life-course realities and perceptions. In some contexts, older persons are likely to consist of those with grown-up children or grandchildren or who are leading figures in societies and other domains. Certain subpopulations, due to adverse life conditions, may face biological signs of ageing earlier than others – such as refugees or prisoners. Persons who have endured conditions of war, conflict and natural disaster cannot be regarded with the healthy ageing metrics of affluent societies.16 37. Data on older persons, where it is collected, often portrays a homogenous group, whereas it is the most diverse of all age groups, in terms of chronological age, health status, life stages, living arrangements, and needs. Statistical systems need, therefore, to embody a more nuanced understanding of old age and the different forms of participation and potential for contributions by older persons. 38. There are also methodological limitations that constrain the availability and validity of data on older persons. Most surveys, even those dedicated to older persons, fail to take account of older persons in institutional care facilities, older prisoners, or older homeless persons. Among the reasons for this are difficulties in enlisting older persons to participate, due to incomplete official registers of information, or a lack thereof, the use of traditional household surveys, and the use of so-called gatekeepers who have to give their consent. In addition, some populations, such as older migrants, older persons who receive care at home, older persons with dementia, and those living in rural or remote areas, may be excluded due to language barriers, lack of administrative records, physical or cognitive impairments, and remoteness. The use of proxies to overcome these limitations may not yield accurate results. 39. Applying a participatory approach can help to improve response rates among certain groups of older persons. This may be particularly relevant for older persons who experience multiple forms of discrimination, as well as those who are excluded from traditional household surveys, which tend to perpetuate the invisibility of older persons, or those who 14 15 16 A/HRC/42/43, para. 46. Principles and Recommendations for Population and Housing Censuses (2017), revision 3, available at https://unstats.un.org/unsd/demographic-social/Standards-and-Methods/files/Principles_ and_Recommendations/Population-and-Housing-Censuses/Series_M67rev3-E.pdf. Médecins sans frontières, “Older people in crises: A review of MSF’s approach to vulnerability and needs”, available at www.msf.org.uk/sites/uk/files/older_people_in_crisis_final_oct_2012.pdf, p. 4. 9