over their lifetime, rather than the six or seven if they had married five years earlier. Given the
short life expectancy in pre-industrial Europe of about 35-40 years, this restrained level of
fertility diminished substantially the resulting rate of population growth, except at frontiers of
settlement where labour was scarce, land abundant, and marriage consequently early.
Heckscher (1963) thought Malthus’s framework was relevant to Sweden. With the
Swedish church’s good records of marriages, births and deaths, and the Swedish king’s need to
estimate crop yields (for the purposes of taxation), annual time series for Sweden after 1720
appear accurate and show a positive covariation in marriage and fertility with good crop years,
and shortfalls in marriage and subsequently fertility following poor crop years. Temperature and
rainfall data available for Sweden after 1750 allow later analysts to incorporate this exogenous
variation in weather and employ vector autoregression to estimate weather-driven Malthusian
cycles in wages, fertility, as well as mortality (Eckstein, Schultz and Wolpin, 1984).
Working with French and Swiss parish registries of marriage, births, and deaths, Louis
Henry (1972), the demographer, found evidence that couples exhibited a ‘natural’ rate of
childbearing after marriage, until they eventually began to increase the intervals between their
births after later parities, if economic conditions became less favourable. The emergence of this
form of parity-specific application of birth control over the life cycle of marriages was interpreted
by Coale (1973) as an indicator of the onset of the ‘demographic transition’, when cultural
restraints on fertility evolved from ‘natural’ proximate determinants to controlled ‘modern’
reproductive behaviour relying primarily on birth control.
Parish registries were then sampled from England from 1541 to 1871 by Wrigley and
Schofield (1981) to further investigate the Malthusian framework. Lee (1981) found that
increases in marriage and birth rates were related to good weather and resulting declines in the
price of wheat, as Malthus would have expected. But only about half of the covariation in
weather/prices and annual birth rates is due to the fluctuations in first births that follow in the
wake of variations in marriage. The other half is explained by variation in the length of inter-birth
intervals. The latter finding casts doubt on Malthus’s view that in this pre-industrial period
couples did not exercise fertility choices within marriage. This spacing of births in response to
economic wage cycles implied that the adoption of parity-specific birth control may not have
been a cultural innovation, as assumed by Coale, but a customary form of individual behaviour
adopted when additional births were unwanted. Some couples in pre-industrial societies appear