Overall, the household estimates, household projections and communal establishment estimates and the 2011 Census estimates are close to each other.
The 2001-adjusted household estimates for 2011, published prior to the release of the 2011 Census, were within one per cent of the 2011 Census estimates, both for Scotland as a whole and for the majority of local authorities. The percentage change in household numbers between 2001 and 2011, from the 2001-adjusted household estimates, was also within one percentage point of the change between the 2001 and 2011 Censuses, for Scotland and the majority of local authorities. Revised estimates, published in 'Estimates of Households and Dwellings in Scotland 2012', that incorporate the 2011 Census are within 0.4 per cent of the 2011 Census estimates for all local authorities, and part of the difference between the revised estimates and the census is likely to be because of the three months between the March census estimates and the June household estimates.
Estimates of average household size have been published for the first time in 'Estimates of Households and Dwellings in Scotland 2012'. These estimates are close to the census estimates, both for Scotland and its local authorities, although the household estimates tended to underestimate average household size in 2011. The estimates of average household size are likely to have been affected by being based on mid-year population estimates which have not yet been revised to take into account the 2011 Census.
The greatest difference from the census is in the projections of household type from the 2010-based household projections. The household projections over-projected the increases in the proportion of smaller households and the decreases in the proportion of larger households. The household projections are based on projecting the change between the 1991 and 2011 Censuses forward. However, the 2011 Census suggests that between 2001 and 2011, the change in the types of households that people live in were not as great as between the 1991 and 2001 Censuses. This change in trend caused the overprojection seen in the 2010-based household projections. The alternative headship projection, published alongside the principal household projection, produced better projections of household type than the principal projection because it assumed slower changes in headship rates than the principal projection. The household projections methodology will be revised to incorporate 2011 Census and a methodology that incorporates more than two census points will be considered, in order to capture more long-term changes in trends.