Births and Deaths Rates: breaks in series circa 2011

Caused by the rebasing of the population estimates to use the results of the 2011 Census

What has happened?

The population estimates for mid-2002 to mid-2011 that were used to calculate the previously - published birth rates and death rates for those years were based on the results of the 2001 Census. The number of people of each age and sex from the 2001 Census was ‘rolled forward’ to take account of ageing, of births and deaths that occurred after the 2001 Census, and of estimates of subsequent migration to, from and within Scotland. The latest population estimates based on the 2001 Census were for mid-2011, and were published in May 2012.

On 8 August 2013, National Records of Scotland (NRS) published population estimates for mid-2012, together with rebased estimates for mid-2011, that were based on the results of the 2011 Census and subsequent ageing, births, deaths and migration. Rebased population estimates for mid-2002 to mid-2010 were published on 17 December 2013.

The rebased estimate of the total population of Scotland in mid-2011 (which was published in August 2013) was about 0.9% higher than the original estimate for mid-2011 (which was published in May 2012). Therefore, there was a break in the time-series of NRS's estimates of the size of the population, until the rebased population estimates for mid-2002 to mid-2010 were published in December 2013. It follows that, for several months, there was a break in every time-series of birth rates and death rates that was produced using the mid-year population estimates.

There were breaks in most of the Vital Events statistical series between August 2013 and 19 February 2014, when NRS published revised versions of most of the relevant Reference Tables for 2012 plus many new Time-Series Tables. (The breaks in the age-standardised death rates had been removed a little earlier, as NRS was able to publish revised versions of those tables on 17 December 2013.)

In general, the revised figures should differ only slightly (if at all) from those which were published previously. A comparison of the 'original' and 'revised' figures for 2011 in several tables found that, in general, the rebasing of the population had reduced the rates for 2011 by at most 1%. However, there were exceptions: sometimes the fall was bigger, and occasionally it was around 3%. There were also cases where the rebasing had increased a rate, and the rise could be as much as about 1%. The extent of the rebasing, and hence of the revision to the birth or death rates, varied with age-group and sex.

The rebasing had more effect on the estimated size of the population of Scotland for 2011 than for any earlier year, and had relatively little effect for 2002 to 2006. It is unlikely that the rebasing has changed significantly the direction or 'shape' of any trend in birth or death rates, because the size of the revision to the estimate of the total population has varied gradually from one year to the next - so the scale of the changes to the birth and death rates should vary gradually from one year to the next.

Taken together, the revised Reference Tables and the new Time-Series Tables cover all the figures which, it is thought, users of these statistics are likely to require for 2002 onwards. However, should any other revised figures be needed, NRS will consider requests from users for revised versions of any of the 'single year' tables for 2002 to 2011: NRS would produce them using the rebased population estimates for those years. However, NRS cannot guarantee to produce such tables: it will depend upon the staff time available, the amount of work involved, and the priority of other tasks. Requests for figures which cannot be obtained from the tables that have been published so far should be sent to Statistics Customer Services.
E-mail: [email protected]