Foreign-born population now 10 million

Fertility

In the UK, vital events (births and deaths) are not registered according to ethnic
origin, only by the birthplace of the mother or of the deceased. The ethnic origin of
each birth in NHS hospitals is noted in an NHS database: by 2005 89% of births were
attributed to an ethnic category (Moser et al. 2008). The ONS Longitudinal Study
links a one per cent sample of the census population to births, deaths and migration.
But ethnic sample sizes are relatively small. Indirect approaches have related the
ethnic group of each child in the 2001 census to the ethnic group of their co-resident
mother (Large and Ghosh 2006; Large and Ghosh 2006), or using the ‘own-child’
method through the Labour Force Survey (Coleman and Dubuc, 2010). These
methods, discussed in the latter paper, do not give identical results. The own-child
method relates children to their own mothers in the same household or family group.
Results from several successive surveys provide estimates of the births in the same
calendar year (subject to the constraints of the panel design of that survey),
substantially improving sample size. That method has been used to estimate fertility
indirectly in studies where direct data are not available (Cho, Retherford et al. 1986)
(Brown 1982; Berthoud 2001), and was used here.

In order to make ethnic projections, assumptions need to be made on the future
development of the fertility of each ethnic group. These were based upon the timeseries
trends from the 1960s and 1970s reported in an earlier paper (Coleman and
Dubuc 2010), adjusted for recent developments as noted below. In some groups
(Black Caribbean, Other Asian), annual total fertility has fallen to about the level of
the national average and has shown little recent trend. Among others (Indian,
Chinese) fertility has fallen below, and remained below, the national average. Only
Bangladeshi and Pakistani women retain fertility substantially above the national
average, albeit with a declining trend. Fertility of Black African women, and that of
the heterogeneous ‘Other‘ category, remain somewhat above average with an
uncertain trend. All show a marked tendency towards delay in childbearing.

Continued decline was assumed among those populations where fertility is currently
elevated, not otherwise. Simple model curves (exponential, logarithmic, logistic)
fitted to Pakistani and Bangladeshi fertility trends from the 1970s up to 2005 pointed
to a decline to 2.0 variously between 2010 and 2040. The estimates adopted are given
in Table 3. It was not assumed that fertility would converge to a uniform level. A
persistently depressed socio-economic and educational position, low engagement in
the workforce among Bangladeshi and Pakistani women, specific religious and ethnic
influences, may take a long time to change. Continued immigration from countries
with incomplete fertility transitions may also retard fertility decline. In similar
projections for other Western countries, fertility of populations from less developed
areas is assumed to stabilise at slightly above the national average e.g. (Statistics
Norway 2008, Statistics Sweden 2008) or to decline only slowly (US Census Bureau
2008). Very likely the delay in childbearing noted in all groups has depressed period
total fertility in UK ethnic groups, as in the US (Yang and Morgan 2003). If so some
recovery to a higher level may be expected. However lack of appropriate data
prevents calculation of a tempo-adjusted total fertility for UK ethnic groups.

The increase in UK total fertility to 1.95 (2008) needs to be taken into account.
Recent immigration has increased the number of immigrant women. Births to
immigrant women comprised 24% of all births in England and Wales in 2008
compared with 13% in 1997, and 58% of the increase in births in England and Wales
between 2002 and 2008. Despite that, data from the Labour Force Survey show that
the actual fertility rates of overseas-born women remained level between 2004 and
2007, although overall, they continue to have higher average fertility than UK-born
women: (2.51 compared with 1.79; Tromans et al. 2009). Fertility of UK born women
(including UK born members of ethnic groups), however, increased from 1.68 to 1.79
over the same period (ONS 2008a). The total fertility estimates derived from earlier
LFS therefore need up-rating.

Following the 2008-9 economic imbroglio and rising unemployment, it is assumed
that this overall fertility increase will cease. In all the projections presented here,
overall UK total fertility is assumed to remain at 1.95 in 2009 and 2010. That gives a
total fertility for 2006-11 of 1.91, a little lower than the assumption in the ONS 2008-
based projection (1.93). For comparability with the ONS Principal Projection, the
OXPOP ‘standard’ projection will assume that long-term overall UK fertility will
revert after 2015 to the ONS long-term assumption of 1.84.

The fertility rates of individual ethnic minority groups are likely to have shared in the
general upswing up to 2008, not the least because of the substantial recent inflow of
immigrant women, with their higher birth-rates, into those groups. In the higher
fertility groups such as Pakistanis and Bangladeshis, no increase is assumed. Instead,
the expected decline in fertity is assumed to have stopped for five years. Otherwise, it
is assumed that all groups share in this increase and a subsequent reversion to a lower
level (Table 3).