Northern Ireland census 2021 – untangling the data

Dr Joe Frey, CaCHE Knowledge Exchange Broker, Northern Ireland, reflects on the recent publication of phase two of the Northern Ireland Statistics and Research Agency’s Census 2021 statistics in December 2022.

The housing-related statistics included as part of this release provide an opportunity to rebase our understanding of Northern Ireland’s housing market and the demographic trends that are essential in shaping it.

Arriving at a welcome time as, due to the impact of the Covid emergency, the Northern Ireland Housing Executive has so far been unable to undertake its quinquennial House Condition Survey -this statistical source is seen as providing the most comprehensive view of housing in NI.

An ageing population

The first main release from Census 2021 (in September 2022) confirmed one of the most significant ongoing demographic trends characterising the advanced economies of western Europe: an ageing demographic profile. In 2011, the proportion of NI’s population aged 65 and over was 14.56%.

By 2021, this had increased to 17.15%, representing an absolute and proportional growth that has significant implications for both the construction of new housing and the refurbishment of existing stock, as well as the housing support services needed to enable older people to continue to live in their own home with an appropriate degree of independence.

Household projections

The 2021 census figures confirmed the robustness of the population projections produced by NISRA on a biennial basis. NI’s 2016-based population estimate (principal forecast) overestimated the actual 2021 population by a mere 1,027 (0.05%). Indeed, even the 2016-based estimate of the proportion of the population 65 and over was remarkably accurate (17.26%) using the zero net migration variant, compared to an actual 17.15%.

However, from the planning for housing perspective, household projections play a more critical role but are considerably more challenging to forecast. It is well known that the rate of household formation is much more heavily influenced by changing economic circumstances and government policy decisions than just the rate of population increase. The most recent (2016-based) household projections estimated that there would be 744,754 households in NI in 2021. The 2021 Census showed that there were 768,810 – an additional 24,056 (3.2%).

This discrepancy is understandable given that the 2016-based projection uses a two-point exponential model that is based on the demographic trends that prevailed during the 2001-2011 decade – a decade characterised by suppressed household formation due in the early years of the decade to a growing affordability crisis and in the latter years the severe economic downturn that followed the Global Financial Crisis of 2007/08.

The discrepancy has significant implications for planning for housing. The most recent Housing Growth Indicators (2016-2030) published by the Department for Infrastructure and the Housing Executive’s more recent estimate of new dwelling requirements (2020-2035) are ultimately based on the demographic trends that characterised the period 2001-2011.

There is little doubt that when these estimates of future housing need and demand are updated based on the Census 2021 statistics, more ambitious housing supply targets will need to be set in order to address NI’s well-recognised undersupply of housing (social and private).

Private rented sector

Perhaps the most important housing statistic to emerge from the 2021 census is the size of NI’s private rented sector (PRS). The census confirmed that a total of approximately 150,000 households (19.5% of all households) were living in the PRS and that there had been no significant change in the proportion of households living in the sector over the previous five years (when the 2016 NI House Condition Survey (NIHCS) provided a robust estimate of the overall size of the sector: 18.3% of households).

For me, this answers the doubts raised in 2022 by two of the leading household surveys in NI (the Continuous Household Survey (CHS) and the Family Resources Survey (FRS) that published figures indicating a much smaller proportion of households now living in the PRS (CHS: 14% in 2021/22; FRS 13% in 2020/21). The CHS has for many years significantly underestimated the size of the PRS – mainly because it does not adjust its tenure statistics to allow for the under response of households in the PRS – a response bias that has been shown to prevail for many years by successive House Condition Surveys.

The FRS traditionally provided an estimate of the PRS (19% in both 2018/19 and 2019/20) that was very much in line with census and NIHCS estimates. A reduction to 13% in the space of one year would have implied a reduction of approximately one-third of the total occupied PRS properties – a tectonic shift in the structure of NI’s housing market – something for which there was not even any significant anecdotal evidence, despite some media speculation that landlords would leave the market in droves following changes to taxation and tighter regulation of the sector.

This blog is in no way intended to discredit the tremendous amount of invaluable socio-economic and other data provided by both the CHS and the FRS or the household projections provided by NISRA that underlie the assessment of future housing requirements.

In an increasingly uncertain world, it is important that planners and policymakers take a broader, more informed, increasingly rounded view of data sources and projections, pooling their knowledge and experience in a cross-organisational collaborative approach. This would include harnessing a more comprehensive range of qualitative and quantitative data and insights from the private sector.

As a result, planners and policymakers may feel more empowered to appropriately adjust and moderate essential time series-based statistics to optimise key estimates and targets that underpin housing policy and planning for housing in NI. Census 2021 provides a springboard to achieve this.


Date: February 8, 2023 2:22 pm


Categorised in: , ,