Jim Oeppen

Campop’s estimated series of population totals for England from 1541 to 1871 are the longest and most detailed available for any country. The associated age-structures have been used to provide summary measures of fertility and mortality, such as replacement rates and life expectancy. The opportunity they present for extending per capita analysis into the past means that they have become a standard reference for historical demography and economic history, and have been cited in over 1,500 academic publications. 

Why do we need to calculate population size?  

Expressing data “per capita” is an essential part of how we understand social and economic statistics today, but this kind of calculation relies on knowing the population total.  Before 1860, the term “per capita” was extremely rare.  (Try typing “per capita” or “per person” into Google Books Ngram Viewer.)  

This is not surprising when you consider that there was no census of England and Wales before 1801, and age-breakdowns only began in 1821. Part of the rationale for taking the first census was that politicians couldn’t decide whether the population was increasing or decreasing.  It is surprising that they didn’t know, as we now know that the population was increasing very rapidly! 

So how do we calculate historical population size without a census? 

In the absence of census data before 1801, Campop’s population estimates are derived from annual counts of baptisms and burials. After adjustment, these can be used as proxies for births and deaths.  Official registration of births, deaths, and marriages began in 1837, but the established Church in England had been registering baptisms, marriages, and burials since 1538.   

Ascension parish burial ground Cambridge. Source: Wikimedia.

Campop encouraged their army of volunteers to find local parish registers and count the events in each month. The initial aim was not to estimate the population, but to find registers without gaps that might be suitable for Family Reconstitution (a technique for the detailed analysis of family demography).  

However, once the counts started coming in, it became apparent that they were an important resource in their own right. It was decided that they could be used as a sample that could be inflated to represent national estimates of births, marriages and deaths. 

A total of 404 parishes were identified that satisfied a suite of criteria for accuracy and completeness. These represent a four percent sample of the 10,000 ancient English parishes. They provide 3.7 million baptisms, marriages, and burials in monthly totals.  

Problems and adjustments 

These 404 parishes were not a random sample. The biggest issues were that large parishes were over-represented; there were no London parishes; and the series started and stopped at different dates.  For example, between 1662 and 1811 all 404 parishes were in observation, but in earlier and later years the number declined. 

Going from baptisms and burials to “national” totals of births and deaths involved a long series of adjustments. These included inflating the births to allow for deaths before baptism; and adding under-counted non-Conformist births and events for London. Finally, the sample was re-weighted so that the large parishes lost their dominance in the counts.    

Validating the estimation method 

Sweden has records of deaths by age, censuses, and life tables from 1751.  It also has high levels of net-outmigration in the 19th century.  We can pretend we don’t know about the censuses, and only use the totals of births and deaths with the 1901 census as targets. The estimation method was able to satisfactorily match the observed censuses from 1751 onwards, and the net-migration rates. 

The estimated values for English life expectancy derived from totals of births and deaths are remarkably consistent with those derived from subsequent research on individual life histories in Campop’s English Family Reconstitution studies.  

Figure 1 (below) shows Campop’s estimates of the population of England from 1536 to 1796, together with the decadal Census counts for England from 1801. The period from 1541 to 1651 exhibits rapid exponential growth, followed by a period of stagnation, before exponential growth returns in the 19th century, but with a slower rate after 1901. The population doubling time was about a century in the first growth period, but became even more rapid in the 19th century, shortening to 50 years. 

Findings

Using this information, Campop members were able to make a series of discoveries about populations in the past, including:

Dependency ratio 

A historical definition of the working age population is between the ages of 15 and 60, with those older and younger regarded as ‘dependent’.  Although there have been changes as the population grew and stagnated, and these changes had important social and economic consequences, the Dependency Ratio hovered around 70 dependents per 100 providers from 1551 to 2021. This stability masks an important transition. Before 1900 there were five children for every elderly dependent. The 20th century saw this fall to one child for each elderly person.  

Georgios Iakovidis, The Favourite (1890). Source: Wikimedia

Net migration 

The net-migration estimates show consistent out-migration at the low rate of about one to two persons per thousand per annum. It is likely that in-migrants, particularly from Scotland, Ireland and Wales, almost balanced English emigration to the rest of the world. 

Fertility and life expectancy 

The pre-census population estimates can be used to derive demographic measures of fertility and survival from 1551 to 1841.  For example, we found that women who survived to age 50 only had about five births on average, which is very different from typical expectations about the past – look out for a blog post about this next week! 

Life-expectancy at birth derived from these populations was generally between 30 and 40 years, but around 20-25% of people were aged 40+.  If you wonder how this can be possible, it will be discussed in the forthcoming blog “Three Score and Ten”.

Population dynamics 

From these estimates, we can see that the impact of epidemics on England’s population has been minor and transitory. From 1541 to 1851, changes in fertility were more important than changes in survival in determining population growth.  The importance of age at marriage as a factor influencing fertility is discussed in another of today’s blogs…

Further reading 

• Human Mortality Database. Max Planck Institute for Demographic Research (Germany), University of California, Berkeley (USA), and French Institute for Demographic Studies (France). Available at www.mortality.org
• Lee, R. D., ‘Estimating Series of Vital Rates and Age Structures from Baptisms and Burials: A New Technique with Applications to Pre-industrial England’, Population Studies 28 (1974), 495-512.
• Wrigley, E. A., and Schofield, R. S.,  The Population History of England 1541-1871: a reconstruction (Cambridge University Press, 1981).
• Wrigley, E. A., Davies, R. S., Oeppen, J. E., and Schofield, R. S., English Population History from Family Reconstitution: 1580-1837 (Cambridge University Press, 1997).

Tags: census, demography, parish registers, population size

| This entry was posted on Thursday, July 11th, 2024 at 12:01 am and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.