It's used (historically) because household income is one of the best determinants of living standard for a great majority of individuals under "typical" living

It's used (historically) because household income is one of the best determinants of living standard for a great majority of individuals under "typical" living conditions. I.e. if you have one high income earner, married to a low income earner, supporting one or two no income earners living together in a household, its inaccurate to split those out and consider or calculate those statistics as individuals. Most households historically live a lifestyle more akin to pooling and sharing income and resources. But, as you rightfully point out, when I was calculating them, we also tended to use equivalised-household incomes (to take into account the different resources and structures used by the number of children vs adults living together). 'Course you also then have to worry about things like imputed rent and drawing on access to capital/resources that don't turn up in economic or monetary flows. The short answer for why its used is probably like most economic statistics: its calculable and easily available and people think they understand it (even if they don't).