I’ve just read a fascinating article by William Davies on the history and current use of statistics. Even the etymology of the word, which had never occurred to me, is fascinating: the term originated in Germany in the second half of the seventeenth century as a way for the empire to map variations in culture and laws across its many statelets. It focussed on populations in broad terms rather than on the usual centres of power and wealth.
The use of statistics expanded during the Enlightenment as nations counted births, deaths, marriages, imports and exports, and statistics became linked to the idea of progressive ideals. With new mathematical techniques, it became possible to view highly complex information in an early form of a spreadsheet. The disadvantage has always been that this broad national sweep flushes away regional or cultural differences, which is one reason for public dislike and mistrust of them. Statistics may work best with static, homogeneous populations where ironed-out differences are not so distorting.
So why are statistics and experts so derided nowadays?
Naturally, statistics are not objective: who decides what is measured? For example, GDP only recognises paid work, so traditional “women’s work” slips through the net; France has not collected census data on ethnicity since 1978, which is both egalitarian and makes it impossible to quantify employment racism in action. But statistics do make it possible to identify improvements or deteriorations over time in areas like public health or industrial efficiency.
Later collection of statistics expanded from centralised government to early social scientists like Charles Booth as part of his inquiry into urban poverty. Opinion polling is another form of statistical survey; the author thinks that our obsession with it is part of our belief in mass democracy. He sees the recent rubbishing of statistics as a reflection of public hostility to the view of society as a whole or the economy as a whole; the cry from the individual is What about me? Where is the place for collective memory or a sense of place in the statistical average? The Eurozone illustrates this sense of dislocation: it incorporates about half a billion people and several very different countries, but their individual fates are splintering into very different forms.
The other problem with statistics at a micro level is that they don’t measure intensity: Davies here refers to employment. For government purposes you are either employed or seeking employment; only recently has the ONS considered underemployment – a problem which affects about 6% of the workforce. Nor are statistical classifications flexible enough for individuals who cleave to identity politics.
Finally, Davies looks at statistical data which is collected automatically (rather than via government’s desire to answer a specific question) through digital activity. Facebook, for example, has the ability to carry out enormous quantitative social science through the data it gathers from its users, but it has no reason to do so for anything other than profit. (Although I do recall listening to something on More or Less about Google searches being interrogated to track the spread of some kind of vomiting bug.) “Sentiment analysis” can track public moods by focusing on key words used on social media, and Donald Trump’s campaign made use of this data.
But a modern world without good quality statistical data would be horrendous; they may need tweaking, but they are still crucial for a liberal and enlightened society.