Inside Statistics: The double-edged knife of statistics
A must read for any chief statistician is an article by Wayne Smith, the former chief statistician of Canada, penned in 2018, titled, Canada’s New Statistics Act.
This is so because I had the benefit of inheriting a modern Statistics Law from Mark Orkin, my predecessor at Statistics SA.
I had the prospect of making improvements and amendments on this, but my time was up.
I have left my successor, Statistician-General Risenga Maluleke, with the burden of pushing through the reforms I was eager to implement as a handover note to him.
On the 21st of February 2011, I was invited to address a topic on the Threats to the United Nations Fundamental Principles of Official Statistics, (Unfops), during the 42nd Session of the United Nations Statistical Commission in New York.
The series of seminars was a build up towards adopting these principles by the United Nations General Assembly. In the mix of threats to the principles were content areas on the most sensitive of statistical series.
These included the consumer price index, the most sensitive as it has power of swaying the amorphous markets, the gross domestic product, the census and government debt.
The offenders were a mixed bag, but largely politicians and chief statisticians.
The countries were a mixed bag, ranging from the habitual offenders, the usual suspects and strangely the angels were in the bag too.
The effects of the violations were not constrained to country boundaries, but extended beyond and fuelled major political tensions.
Amongst the habitual offenders was Argentina where the political authorities interfered with the production and level of the Consumer Price Index and presented a figure that was much lower than what the staff in the statistics office produced.
This ultimately cost Graciella Bavaqua her job as head of prices at the statistics office because she refused to accept the government bullying and interfering in the work of the statistician-general.
Greece came fast on the heels of Argentina, but here it was a violation on the measurement of government debt.
Andreas Georgiou, the head of Elstat corrected and revised the debt upwards to the displeasure of the Greek government.
Georgiou ten years later is still in and out of the Greek courts because the government has labelled him enemy number one as Greece reeled from an unending economic crisis.
As regards South Africa, I had to lay bare the linen on the 2002/2003 exaggerated Consumer Price Index that ran for 15 months and caused untold suffering,especially for pensioners who bought inflation linked bonds who lost on their investment, when I revised the Consumer Price Index downwards.
Indonesia had been seduced into providing individual records from the census to assist the government in addressing poverty and Rusman Heriawan was now in trouble as the government demanded new lists for the poor. This violated protection of privacy.
Israel was on the allegation box because Palestine accused them of attacking its statistics office and making away with the census records in order to execute targeted arrests and bombing of Palestinians and their houses.
This violation of privacy of records is against the fundamental principles.
Canada’s case was about the minister changing the way the census questionnaire was deployed and making the mandatory answering of questions optional.
Wayne Smith is raising this particular point of political interference in the professional activities of the chief statistician.
What is puzzling, but redeeming though is Smith’s about turn. In that Seminar in New York six years ago he objected strongly to this aspect of my presentation as a fundamental methodological violation of the fundamental principles.
He was quite measured, however, compared to the Argentinians who hurled insults at me before they stormed out of the meeting.
Later that evening they raised a diplomatic protest against me to Ambassador Baso Sanqu, the representative of the Permanent Mission of South Africa to the UN. I was summoned, but the ambassador was satisfied with my position.
What is important however, is this – both Argentinian authorities and Smith have now turned the corner and come to accept that these were major violations. They are on the mend.
The South African Statistics Act was way ahead of Canada, which depended mostly on convention.
Smith in his article argues that convention is just not enough and both laws of Argentina and Canada have been put under the knife for major surgery.
Dr Pali Lehohla is the former Statistician-General of South Africa and former head of Statistics South Africa
BUSINESS REPORT ONLINE