OPINION11 September 2013

Census: Why the focus on ‘progress’ is problematic

Opinion

Northstar’s Peter Zoutis warns that replacing the 200-year-old UK Census would constitute one of the greatest misfortunes of modern time.

Res_4010443_Peter_Zoutis_458

According to Peter Zoutis, a principal at Northstar Research Partners, the latter option would constitute “one of the greatest misfortunes of modern time”. He explains:

Res_4010443_open_quote

The national census has been and continues to be the only real marker for understanding society’s base level of existence – its demography and subsistence characteristics. A decision to eliminate the census entirely or supplement it by using alternative information sources would most certainly compromise the integrity of the only existing population counter and descriptor.

Without doubt social research has come a long way since the start of the 19th century, especially in the past 25 years where everything has gotten faster and cheaper year on year, but this focus on “progress” isn’t without its problems. There is much that can be said for maintaining tradition in the area of population measurement and the following three tangible outcomes are key reasons for maintaining the current approach:

  1. Accuracy and truth
  2. Societal utility
  3. Inclusivity and membership

When we think about all the social research conducted in the UK (and many continue to dispute its stature as scientific) the census is the one that stands out for accuracy and truth – there is no greater truth about a society than when you ask each member for his/her participation. The result is an actual population parameter of its general make-up, not merely an estimate from a sample drawn from it. No margins of error, just truth. Just as we know that acceleration resulting from the earth’s gravity is 9.8m/s2 exactly, we know the population’s composition precisely – no ifs, ands or buts about it – and that is a very powerful claim to be able to make.

Another benefit to the census is its societal utility. Aside from its ability to describe the population on certain characteristics that will aid in the development of public policies, it’s the one “survey” used by many, if not all, social research organisations to benchmark their data in the pursuit of a nationally representative sample – one fundamental principle to enhancing the external validity of a sample.

Finally, above and beyond all things mathematical and statistical, the census offers inclusivity and membership as every UK resident is asked to participate. In no other way can everyone express their “membership”. It is both a right and a privilege of each resident to complete a census all the while knowing that each is a member of that “club” contributing equally and without discrimination (richer or poorer, upper class or working class, British origin or non-British origin, younger or older, property owner or renter, etc.).

Res_4010443_close_quote

Peter Zoutis is a principal at Northstar Research Partners

@RESEARCH LIVE

3 Comments

11 years ago | 1 like

I never agreed more strongly with any article in Research News than this one. I live in Australia where censuses are taken (amazingly) every five years. The thought of the UK having no census at all, where the British census has been the model for so many others, is incomprehensible. Hopefully the market and social research community will fight this suggestion.

Like Report

11 years ago | 1 like

The article by Peter Zoulis describes exactly and precisely the need for the traditional Census in this country to continue. By all means use intermediate data, but the "right and privilege of each resident to complete a census" must be maintained. It is incomprehensible why anyone could think otherwise.

Like Report

11 years ago

It is misleading to consider any national census as the truth. Census data comes with confidence intervals and caveats. Anyone that has worked with survey data knows the many issues associated with it. I'm sure there is much information lost when processing the census forms and converting them into statistics. At least with a survey we are more cautious interpreting it.

Like Report