Populate is an population projection program for historians.
It uses minimal data (births, deaths and total migrants if known)
to compute sophisticated statistics (life expectancy, age structure of
the population, refined fertility statistics and so on) at annual or
quinquennial intervals over very long periods of time (from a few
decades to several centuries, up to 640 years, if you have the data or
are willing to invent the data).
The inverse method contradicts common sense. Instead of
beginning in the present with the best data and working back in time,
it begins in the past (often with arbitrary initial states) and works
forward toward the present, even into the future. As it turns out,
for projection purposes, it is easier to project forward--that is,
insert births at the bottom of the population pyramid and subtract out
deaths along the slope--than to project backward--add deaths along the
slope and subtract out births at the base.
(The biggest problem for back projection occurs at the very top--
small mistakes here grow as each cohort steps down the pyramid at each
projection, small errors being magnified at each cycle. Note however
that the Cambridge Group's reconstruction of England's population from
1541 to 1870 was done with back projection. Note as well that the
back projection program has not been released for use on a
microcomputer.)
Inverse projection has a doggerel: "vital rates, not initial
states". That is, population pyramids are determined by vital rates
and rapidly forget any initial state. Thus, a projection may begin
with an arbitrary age and mortality structure, because true ones will
quickly take hold if the annual vital rates are correct. Don't
believe it? Check the proofs in my essays on Norway, Sweden and a
"double-blind" challenge with a demographer using simulated data
(McCaa and Vaupel, 1992).
Ron Lee's 1974 essay in Population Studies explains the method
and applies it to Colyton. My essays on Norway and Sweden demonstrate
the method's uncanny ability to elicit to a small degree of error a
broad range of demographic parameters from minimal inputs.
The file biblio.asc lists a more complete bibliography.
Why use populate? For teaching it allows students to do "what
if's"--to find out how demographics work. Consider the old chestnut:
"What caused the modern rise of population in England?" The
conventional answer can easily be demonstrated to be in-correct. How?
Run populate. Select England at the first screen. After the first
projection, select Learn and follow the lessons (this is explained
further in the manual--populate.asc). The exercise consists of two
additional steps: holding fertility constant and then holding
mortality constant. By comparing the results of the first projection
(what actually happened) with the simulations, the effects of changing
mortality and fertility can be readily appreciated.
Or consider Mexico. What would Mexico look like demographically
in say 2020, if from 1930, it had followed the demographics of Chile?
or Spain? or if fertility decline had begun in 1950-54, instead of
1970-74? To find out, run populate and access the data file
"mexico.pop".
For researchers, Populate can be used to develop a consistent,
detailed population balance sheet, using whatever facts are available
to elicit population statistics which would otherwise remain unknown.
To create your own dataset for analysis, use one of the ".pop" files
as a guide or see the populate booklet, "populate.asc".
Lee, Ronald L. "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.
McCaa, Robert. "Populate: A Microcomputer Projection Package
for Aggregative Data Applied to Norway, 1736-1970", Annales de
D:mographie Historique, 1989:287-298.
-----. "Benchmarks for a New Inverse Population Projection
Program: England, Sweden, and a Standard Demographic Transition," in
Roger Schofield and David Reher, eds., Old and New Methods in
Historical Demography (Oxford: Oxford University Press, 1993).
rmccaa@vm1.spcs.umn.edu