previous | contents | next

Household Projections for Scotland, 2010-based

4. Variant Household Projections

4.1 Introduction to the variant household projections

4.1.1 Variant projections provide a broad indication of the sensitivity of the household projections to the demographic assumptions that are used to produce them. The principal household projection is based on the principal population projection produced by National Records of Scotland (NRS), which uses assumptions about fertility, mortality and migration which are thought most likely to occur over the next 25 years. The household projection is also dependent on the assumption that the trends in household formation that occurred between the 1991 and 2001 censuses will continue until 2035.

4.1.2 Three types of variant household projections are presented here. The first set of variant projections is based upon variant population projections produced by NRS that use different assumptions regarding future migration (Section 4.2). The second set of variant projections use different assumptions regarding future trends in household formation, and combine these alternative assumptions with the principal and variant population projections described above (Section 4.3). The last variant produced investigates the possible effect of limited household formation that may occur if the current economic downturn continues (Section 4.4).

4.1.3 This publication presents summary results for each variant projection. Table 20 shows the total number of households projected for 2035, for each variant, by local authority. Table 21 shows the proportion of households in each household type in 2035, for each variant. These, and additional comparisons, are discussed below. Full results of each variant projection, by local authority, are available on request.

4.2 Variation in migration assumptions

4.2.1 NRS produces variant population projections that are based on alternative assumptions of future fertility, mortality and migration. The low and high migration variant population projections use the same assumptions about fertility and mortality as the principal population projections but assume varied levels of migration to and from Scotland. The migration assumptions, by local authority, can be found in Annex E and F of the '2010-based population projections for Scottish Areas'.

4.2.2 Migration variant household projections are calculated by replacing the principal population projections with the variant migration population projections, keeping all other inputs to the household projections (communal establishment rates, headship rates and household estimates) unchanged.

4.2.3 The principal household projection shows the number of households in Scotland increasing from 2.36 million households in 2010 to 2.89 million households in 2035. This compares to 2.77 million households for the low migration variant and 3.00 million households for the high migration variant in 2035. Figure 8 illustrates this, along with the projected increase in population for the principal and variant projections. The percentage increase in households between 2010 and 2035 is 18 per cent for the low migration variant and 27 per cent for the high migration variant, compared to 23 per cent for the principal projection.

Figure 8: Principal, low and high migration variants, 2010-based population and household projections for Scotland

Figure 8: Principal, low and high migration variants, 2010-based population and household projections for Scotland

4.2.4 The use of different population projections has little effect on the type of households that the population is living in. In 2035 average household size in the low and high migration projections is 1.94 and 1.96 respectively, compared to 1.95 in the principal household projection and the proportions of the population in each household type differ from the principal projection by less than 0.3 percentage points (Table 21). The differences are small as the assumptions about household formation are the same in each projection. The small differences in each projection are due to differences in the age distribution of the underlying populations, as different age groups tend to form different types of household.

4.2.5 There are differences in the age groups of the heads of household between the projections. These are caused by different age distributions in the variant population projections, as the age groups of migrants are not necessarily typical of the overall population. The differences in the proportion of all households that are headed by each age group, compared to the principal population, are small (less than one percentage point) with both variant projections (Table 22). The largest differences are seen in the 30-44 age group, where the proportion of all households is 0.9 percentage points smaller than in the principal projection in the low migration variant and 0.8 percentage points larger in the high migration variant.

4.2.6 Table 23 shows the projected number of households for each local authority under the principal, low and high migration variants for 2035. The low migration variant has the biggest impact on the City of Edinburgh (where the number of households is seven per cent lower than the principal projection) and the smallest impact on the Orkney Islands (one per cent lower). The high migration variant has the biggest impact on the cities of Aberdeen and Edinburgh (where the number of households is seven per cent higher than under the principal projection). In 12 local authorities the high migration variant only has a small impact (increases of around two per cent).

4.2.7 Figure 9 compares the percentage change in households from 2010 to 2035 under the principal, low and high migration variant projections for each local authority. The impact of low and high migration variants on different local authorities can be seen here.

Figure 9: Projected percentage change in households, 2010 to 2035, principal, low and high migration variant projections, by local authority

Figure 9: Projected percentage change in households, 2010 to 2035, principal, low and high migration variant projections, by local authority

4.3 Variation in household formation assumptions

4.3.1 The household formation information used in the household projections is in the form of headship rates, derived from the 1991 and 2001 censuses. Headship rates describe, for each age group, the proportion of the population that heads a household. In the principal projection, the headship rates from the 1991 and 2001 censuses are projected forward to 2035 using a modified exponential equation (described in Section 5.4).

4.3.2 For the first time, this section presents household projections using an alternative assumption about future trends in headship rate. The alternative assumption used is that, from 2002 onwards, the headship rate changes at half the rate given by the principal projection. This slower rate of change was chosen as results from the Scottish Household Survey suggest that fewer people in Scotland are tending to live in smaller and smaller households compared to the principal household projection. Further discussion is available in the Household Analysis Review Group papers HARG 2011(4) and HARG 2012(5).

4.3.3 Alternative headship variant projections are calculated by replacing the principal headship rate projection with the alternative headship rate projection described above. In the main alternative headship rate projection all other inputs to the household projection (population projection, communal establishment rates and household estimates) remain the same as in the principal household projection. The results of this projection are presented in paragraphs 4.3.4 to 4.3.8.

Figure 10: Principal and alternative headship variant household projections for Scotland, 2010 to 2035

Figure 10: Principal and alternative headship variant household projections for Scotland, 2010 to 2035

4.3.4 Figure 10 compares the alternative headship projection to the principal projection. The principal projection shows the number of households in Scotland increasing from 2.36 million in 2010 to 2.89 million in 2035. This is higher than the alternative headship variant which projects 2.81 million households in 2035.

4.3.5 Both the alternative headship variant and the low migration variant discussed at Section 4.2 project lower household numbers than the principal household projection, but for different reasons. In the low migration variant there are fewer households projected because the population is smaller, however this is not the case with the alternative headship projection, as it uses the same population as the principal household projection. The difference is, instead, due to differences in projected household formation, as the headship rates change more slowly in the alternative headship rate projection. Throughout the projection period the average household size in the principal projection decreases from 2.17 in 2010 to 1.95 in 2035, and to 1.94 in the low migration variant, whereas in the alternative headship variant the average household size projected for 2035 is higher, at 2.00 people per household.

4.3.6 The difference in the average household size is also reflected in the proportions of households of each type in the different projections (Table 21). This can be seen in Figure 11, for the principal and alternative headship rate projection. In 2035, the alternative headship variant has a similar proportion of households to the principal projection in two adult only households, but a lower proportion in one adult households, with or without children, and higher in two adult households with children, and in three or more adult only households.

4.3.7 Using the alternative headship rates also affects the age distribution of the heads of household (Table 22). Most of the differences in the proportions in each age group are small. There are decreases of between 0.4 and 1.1 percentage points in the younger age groups (16 to 59) and increases in the older age groups (60 to 84) of between 0.8 and 1.7 percentage points. These differences reflect the fact that older people tend to live in smaller households, rather than households with children or multi adult households. There is very little difference between the alternative headship and principal projections for the oldest age group (85+).

Figure 11: Proportion of households in Scotland in each household type, 2035, principal and alternative headship variant projections

Figure 11: Proportion of households in Scotland in each household type, 2035, principal and alternative headship variant projections

4.3.8 Figure 12 shows the differences between the principal household projection and the alternative headship variant for each local authority, by comparing the percentage change in the total number of households between 2010 and 2035. The greatest difference from the principal projection occurs in Glasgow City (five per cent fewer households) and the smallest differences are in Aberdeenshire, Argyll and Bute, East Dunbartonshire, East Renfrewshire and Moray (one per cent fewer households each). These differences can be seen in Table 24.

Figure 12: Projected percentage change in households, 2010 to 2035, principal and alternative headship variant projections, by local authority

Figure 12: Projected percentage change in households, 2010 to 2035, principal and alternative headship variant projections, by local authority

4.3.9 Additional variants, using the alternative headship rate projection and either the high or low migration population projections, were also prepared. These differ from the principal household projection in two respects (headship rate and population), but differ from the main alternative headship variant only in the population used. Figure 13 compares the alternative headship rate variants, using the high and low migration populations, to the principal household projection and the main alternative headship variant.

4.3.10 The alternative headship projection for 2035, using the low migration population projection, is 2.70 million and using the high migration population projection is 2.92 million. The latter is 28,500 higher than the principal household projection (2.89 million households), but not as high as the high migration household projection (3.00 million households). This is because, like the high migration household projection, the population is larger than the principal household projection, but as the headship rates do not change as much, fewer new households are formed. This is reflected in average household size, which at 2035 is 1.95 for the principal projection, 1.96 for the high migration household projection, but 2.02 for the alternative headship variant, with high migration population.

4.3.11 Table 24 shows the projected number of households in 2035 for Scotland, and each local authority, under all three alternative headship variants and compares these to the principal household projection. The differences, at local authority level, from the principal household projection, seen in the alternative headship variant with low population projection range from two to ten per cent decreases. The differences in the alternative headship variant with the high population projection range from a one per cent decrease to a four per cent increase.

Figure 13: Principal and alternative headship rate household projections for Scotland, using variant migration population projections, 2010 to 2035

Figure 13: Principal and alternative headship rate household projections for Scotland, using variant migration population projections, 2010 to 2035

4.4 Variation related to the economic downturn

4.4.1 Since 2008, Scotland has been experiencing an economic downturn that has affected many parts of the economy, including the construction sector. New housing supply, including new builds, refurbishments and conversions, has fallen in each year from 2008-09 to 2010-11 (Footnote 3).
Figure 14 is taken from 'Estimates of Households and Dwellings in Scotland, 2011' and presents the year on year increase in the household estimates for Scotland, between 2001 and 2011. It shows that between 2001 and 2008 the year on year increase in the household estimates was between 16,000 and 22,900 households, but that in 2009 it was only 12,800, decreasing to 10,600 by 2011. Reductions in new house building, combined with constrained mortgage and labour markets are likely to be affecting new household formation which is reflected in these figures (Footnote 4).

4.4.2 As the household formation trends used in the household projections are based on 1991 and 2001 census data, they do not incorporate recent changes, such as those that might result from the economic downturn. The household projections take into account past changes by adjusting the first two years of the projections to the 2010 and 2011 household estimates, and adjusting the projections for 2012 onwards by the same proportions as the 2011 figures. However, the trend in household formation rates is not affected by this.

Figure 14: Annual increase in the number of households in Scotland, 2001 to 2011

Figure 14: Annual increase in the number of households in Scotland, 2001 to 2011

4.4.3 In order to investigate the possible effect of a continued economic downturn, we have produced a variant projection where for the next three years (2012 to 2014) the year on year increase in household numbers for each local authority has been assumed to be the same as the average annual increase in the household estimates between 2009 and 2011. The projections for 2015 onwards are adjusted by the same proportions as the 2014 figures however, beyond 2014, we have assumed that household formation returns to the projected trend.

4.4.4 Figure 15 shows the projected household numbers for Scotland for the constrained housing variant and for the principal projection. The principal household projection shows the number of households in Scotland increasing from 2.36 million in 2010 to 2.89 million in 2035 (a 23 per cent increase), whereas the constrained housing projection shows a smaller increase of 20 per cent, to 2.83 million households by 2035.

Figure 15: Principal and constrained housing variant household projections for Scotland, 2010 to 2035

Figure 15: Principal and constrained housing variant household projections for Scotland, 2010 to 2035

4.4.5 Table 25 shows the projected number of households for each local authority, under the principal and the constrained housing variant projections. The projected percentage change in household numbers between 2010 and 2035, by local authority, is shown in Figure 16. The differences from the principal projection in 2035 ranged from four per cent fewer households in Aberdeen City, City of Edinburgh, Glasgow City and Perth and Kinross to one per cent more households in Eilean Siar and Midlothian. Increases in the projected number of households occur in local authorities where the average annual increase in the household estimates between 2009 and 2011 was greater than the projected annual increases in the principal household projections between 2012 and 2014. For eight local authorities there was less than a 0.5 per cent difference in the projected household numbers in the constrained housing variant as compared to the principal projection (Argyll and Bute, Dumfries and Galloway, East Ayrshire, East Dunbartonshire, East Renfrewshire, North Ayrshire, Shetland Islands and South Ayrshire).

Figure 16: Projected percentage change in households, 2010 to 2035, principal and constrained housing variant projections, by local authority

Figure 16: Projected percentage change in households, 2010 to 2035, principal and constrained housing variant projections, by local authority

 

previous | contents | next