Period and cohort data (cont.)
Cohort data
Cohort data reflect the real life experience of a cohort of people. A cohort is defined by a shared demographic event which occurred during a specified period of time. This will most commonly refer to births during a set period of time, but may also refer to marriages, onset of puberty, migration, first birth etc.
The "Chinese birth cohort of 1920-1924" refers to all people born in China between January 1st 1920 and December 31st 1924.
The "British female marriage cohort of 1984" refers to all women married in Britain between January 1st 1984 and December 31st 1984.
When popular media refers to "baby boomers" they are actually referring to the post-World War II birth cohort in a number of countries, which includes all people born in the approximate period 1946-1964.
Cohort measures have the benefit of telling us about a group of people’s real life experiences. However they are often not feasible, as data collection takes decades. For example, if we wanted to know the fertility experience of a birth cohort, we would need to wait until every member of that cohort had reached the end of their reproductive life span. For women this may take 50 years. If we were looking at the mortality experience of a birth cohort this would take even longer.
For this reason it is usually only possible to calculate historical cohort measures. We may create partial cohort measures, based on the experience of a cohort thus far. However these are necessarily incomplete as we have no idea what will happen to these individuals in the future.
The differences between period and cohort measures are easier to conceptualise graphically. We will now see how this can be done in the form of Lexis diagrams.