Lies, damned lies, and statistics

On 30 September 2022, the Dutch Statistics Agency reported that the average inflation over September amounted to 17,1% (CBS). It surprised me because it didn’t feel that way. The Dutch Press Agency reported that this 17,1% included the October energy rate increases (ANP). Hence, my blog title.

Something similar happened over the last decade: the enormous increases in housing prices were never reflected in the Consumer Price Index (CPI). Only rental home increases were taken into account. Hence, average CPI was underreported given rent regulation in many countries.

My blog title is short for: “There are three kinds of lies: Lies, Damned Lies, and Statistics”. This quote is generally “attributed to Mark Twain, who himself attributed it to British Prime Minister Benjamin Disraeli, who might never have said it in the first place.” (source, Wiki). Note: URLs added by me.

Obviously, there is a formal justification for the above. Nevertheless, the Dutch Statistics Agency also monitors perceived inflation as “Inflation as people perceive it often differs strongly from the officially measured rate: people always seem to think that inflation is higher than the published figure.” (CBS)

This “strong difference” between official and perceived inflation erodes public trust in government numbers. Additional reporting on that “strong” difference will add fuel to the fire (eg, Ipsos-2022).

Indeed, Pew Research Center reported early June 2022 that American public trust in government over the years 1958-2022 is near historic lows (eg, NYT, WaPo). Late September 2022, a similar Dutch finding was reported (eg, Dutch News, NOS, SCP, Statista, UU). Clearly, failed Covid approaches did not help.

Global populist nationalist politicians are trying to leverage on that low public trust, albeit with mixed results. Most recently, they did win in Italy and Sweden. Perhaps, Brazil can soon be added to this list as its president did not lose the first round, against all odds (ie, opinion polling statistics)

There can be various reasons why statistics cannot (always) be trusted, like low sample size, sample coverage errors, pseudo opinions, or respondent errors. An interesting Brexit related example: would a higher voter turnout (eg, 90% rather than 72%) have changed its Leave outcome?

“33,577,342 votes were cast – that’s a turnout of 72.21% of registered voters. Of those votes, 16,141,241 were to remain a member of the EU. Those are the people who voted against Brexit. 17,410,742 were to leave the EU. Those are the people who voted for Brexit.That’s a majority of one million, two hundred and sixty nine thousand, five hundred and one – in favour of leave.”

Quote by Richard Smedley on Quora

Some harsh quotes on statistics:

• “A single death is a tragedy; a million deaths is a statistic.” Joseph Stalin
• “Facts are stubborn things, but statistics are pliable.” Mark Twain
• “Most murders are committed by someone who is known to the victim. In fact, you are most likely to be murdered by a member of your own family on Christmas day.” Mark Haddon
• “Most people use statistics like a drunk man uses a lamppost; more for support than illumination”. Andrew Lang
• “Statistics, likelihoods, and probabilities mean everything to men, nothing to God.” Richelle E. Goodrich

Little Lies (1987) by Fleetwood Mac
band, lyrics, video, Wiki-band, Wiki-song

Tell me lies
Tell me sweet little lies
(Tell me lies, tell me, tell me lies)
Oh, no, no, you can’t disguise
(You can’t disguise, no, you can’t disguise)
Tell me lies
Tell me sweet little lies

Note: all markings (bold, italic, underlining) by LO unless in quotes or stated otherwise.