Beginning in 1980, the US Census Bureau discontinued the use of the term ‘head of family/household’ and replaced these with ‘householder’ and ‘family householder’, as well as no longer by default classifying the husband in a married couple as the ‘reference person’ in its survey. These logical legal implementations make sense in our society which continues to slowly abolish problematic gender norms bit by bit – but how does this work when it comes to data sets?

How is ‘household’ defined?

When it comes to the realm of data, there is, firstly, a logical requirement to determine what a “household” is.  Is everybody at one address a household?  What if two families are collocated?  What about extended families?  Because the definitions are nebulous, criteria ill defined, and actual data sparse, it is usually necessary to make some assumptions.  Typically, data providers will make the definitional assumption that everybody at one address, sharing the last name, is a household.  Clearly, there are situations where individuals will not be flagged as a household member.  Examples include those who are cohabitating, those under common-law marriage, or even those married but without a name adoption.  There are also some ethnicities where the last name has a different ending based upon gender.  To catch these exceptions, some data providers may look at other data fields, such as joint ownership of a home, or similarity of the last name. Now we know how a household is defined, we can, next time, talk about what qualifies someone as ‘head of household’ – the answer may surprise you!