Heterogeneity
Average experience within a group can be a useful indicator, but there are many instances in which the average, and variations on it, are misleading. These occur when the distribution of vital events within a population is uneven.
Thus, life expectancy at birth is a misleading measure of survival in seventeenth-century Europe because deaths occurred with two modes, infancy and late adulthood. This example is well known, and authorities regularly point to the rapid increase in life expectancy among people who survived infancy and early childhood. Irregular distributions - those taking a form other than the normal or bell curve - are often encountered. They signal something called heterogeneity - incompatible or nonproportionate elements. For example, future health is in part a function of prior health, so that some authorities suggest that a cohort divides itself into two modes, one with a high and the other with a low propensity to be sick. Thus, an average figure - such as the average age for retirement, 65 - may misstate cohort experience in such circumstances.Members of the same cohort may have disparate health experiences, in ways just suggested, but the likelihood of disparity is greater still when different cohorts are compared. The health experiences of a cohort are shaped by the circumstances encountered at each stage in life, so that each cohort accumulates a unique health history. In modern advantaged societies these differences remain, but their magnitudes are such that they have little measured effect on basic vital statistics. Thus, life expectancy at a given age varies little in the short run. But in historical populations the disparity may be marked. In seventeenth-century England, life expectancy at birth shifted dramatically from cohort to cohort. The population alive in 1750, for example, included individuals who had survived such disparate events as the plague of 1665 and the as yet incompletely identified forces of the mideighteenth century favoring mortality decline. Sporadic epidemics caused the number of deaths and the death rate to fluctuate, with the result that the number of years lived by each cohort varied between wide boundaries (Wrigley and Schofield 1981). This shifting cohort pattern is a historical rhythm that deserves recognition.
James C. Riley