National Records of Scotland

Preserving the past, Recording the present, Informing the future

Comparison of the 2000-Based, 1998-Based, 1996-Based and 1994-Based Household Projections

Comparison of the 2000-Based, 1998-Based, 1996-Based and 1994-Based Household Projections

As with the 1998-based household projections the current 2000-based figures were produced using information on households from the 1981 and 1991 Census of Population data, together with the latest set of population projections from the National Records of Scotland (NRS). The 1996-based and 1994-based household projections used the 1971 and 1991 census. The method used to project headship rates taken from the censuses was the same for all 4 sets of projections (go to Annex A for further details).

1. Population Projections

Population projections produced by NRS provide one of the main inputs into the household projections. Changes between the 1998-based and 2000-based population projections will have a direct impact on the 2000-based household projections when compared to the 1998-based household projections.

The population projections are based on population estimates for the base year, and this base population is then projected forward using assumptions about births, deaths and migration.

Chart 2: NRS Population Projections (Persons Aged 16 Years and Over)1994-Based, 1996-Based, 1998-Based and 2000-Based
 

chart

For household projections, the relevant population for the formation of households is the adult population aged 16 and over. Chart 2 shows that the 2000-based projected trend in overall population aged 16 and over closely follows the 1998-based population projections for the years they have in common.

The biggest change between the 2000-based and 1998-based projections has been in the assumption about future levels of fertility. For the new projections, a long-term average completed family size of 1.60 is assumed compared with an average completed family size of 1.75 in the previous projections. The continuing decline in fertility in recent years means that it is now highly unlikely that women born since 1975 will achieve the higher levels assumed in the previous projections.

The chart also shows the higher rate of growth projected in 1998-based figures over that projected in the 1996-based figures. This was largely due to changes in the assumption about migration between these two sets of projections. For the 1998-based projections, migration out of Scotland was projected to continue at a loss of 1,000 persons a year from 2001 onwards; a loss of 3,000 persons a year was assumed in the 1996-based projections. For more information about population projections please refer elsewhere on this website.

2. Categories of Household Type

The categories of household type used for the 2000-based household projections are the same as those used for the 1998 and 1996-based figures.

As for the 1998 and 1996-based household projections, the current set of projections has household type defined in terms of the household composition (numbers of adults and children) and the age group of the head of household.

3. 1998-Based Household Projections Compared to 2000 Household Estimate

The 1998-based figures projected the total number of households in Scotland in 2000 to be 2,204,600. The later estimate for that year was 2,203,160 (a difference of 1,400 after rounding). The 1998-based figures did project a higher rate of growth over the 1996-based projections. Despite this discrepancy between the estimate for 2000 and the 1998-based projection for 2000 it is roughly half the discrepancy seen between the 2000 estimate and the 1996-based projection for 2000.

Chart 3 shows the overall trend in total projected households for the 1994, 1996, 1998 and 2000-based projections. Changes in the population projection and changes introduced to the household projection method choice for the 1998-based projections resulted in a higher rate of projected household growth over that projected in 1996. The same projection method used for the 1998 projections has been used for the current 2000-based projections (for more information refer to Annex A). As a result, and since the population projections are similar, there is very little difference in the overall trend in projected growth between 1998-based and 2000-based household projections (Chart 3 below).

Chart 3: 1994-Based, 1996-Based, 1998-Based and 2000-Based Household Projections (Total Households)

chart