Following the media coverage of data published from the EU-SILC survey in news release 182/2025 “EU-SILC 2024: Main dwellings”, the NSO wishes to provide further clarification on methodological updates implemented in the survey which have resulted in some of the secondary indicators from the EU-SILC 2023 and EU-SILC 2024 to no longer be fully comparable with previous years. Consequently, this change necessitates a break-in-time-series in 2023. The purpose of this note is to provide additional technical information regarding the break-in-time-series to ensure accurate interpretation of the results.

The break-in-time-series was required following the most recent Census of Population and Housing, which was conducted in 2021. While the decennial Census plays a crucial role in providing demographic data, it also serves as a vital benchmark for survey-based indicators compiled within social statistics. Before the 2021 Census, it was not possible to calibrate the EU-SILC data using separate weights for the Maltese and foreign populations; therefore, calibration was done using total population weights. Calibration is a statistical adjustment method applied to survey data to primarily address non-response bias, thus ensuring that all population cohorts in the target population are adequately represented in the survey.

While comparing the EU-SILC weighted counts to the Census figures showed that estimates at the total level were aligned, the distribution by broad citizenship category was not. An analysis of the data showed an under-coverage of the foreign population and consequently an over-coverage of the Maltese. After the 2021 Census data became available, it became possible to apply separate calibration weights for the Maltese and foreign populations, starting from reference year 2023. While this ensured that the distribution by broad citizenship category aligned with the Census, it led to a reduction in the share of Maltese households and an increase in the share of foreign households across the total number of households in the survey.

The NSO does acknowledge the possibility that the under-coverage of foreigners has impacted indicators emerging from the EU-SILC and subsequently their comparability with the estimates produced after the update. An analysis was therefore conducted, through several simulations, to re-estimate the indicators to adjust for the under-coverage noted when benchmarking with the Census. This analysis showed that, while the update adjusted the weighted count of foreigners and Maltese in the survey, it did not significantly impact the rates for the headline indicators for the total population. These are the core indicators produced as part of the EU-SILC and include the at-risk-of-poverty and social exclusion rate (AROPE), at-risk-of-poverty rate (ARP), measures of material and social deprivation (MSD) and low work intensity (LWI). As a result, further to consultation with Eurostat, a revision of the time-series was not deemed necessary to ensure full adherence with EU Regulation (2019/1700). Moreover, such a complex and extensive revision requires technical expertise coupled with rigorous research which has, to date, not been concluded.

There were however some secondary indicators where the estimates for the total household population were significantly impacted by the upward revision in the count of foreigners. This includes the distribution of main dwellings by tenure status. This occurred because the distribution of tenure status across the two broad categories of citizenship varies significantly. As noted previously, when the calibration methodology was updated, it increased the number of foreign households (households where all members are foreign), while reducing the number of Maltese households (households where all members are Maltese) in the survey population. Table 1 presents the weighted counts of households by tenure status from the EU-SILC survey from reference year 2020 to 2024. The data is presented for the total households as well as Maltese households, foreign households and mixed households (where at least one member is Maltese, and one member is foreign) separately. The break-in-time-series is flagged in 2023. The shift in the weighted count of Maltese and foreign households can be seen when comparing the weighted count of households for 2022 to 2023 for the two sub-groups of citizenship and the total, across the tenure status categories. This significant shift in the weighted count of households necessitates a break-in-time-series since the weighted counts prior to the methodological change have not been revised.