Many people will probably remember Pali Lehohla for his yellow suit, but the former statistician-general has made an enormous contribution to our society during his 35-year career in the public service.
Lehohla donated the bright suit - which he introduced to whip up a sense of excitement around Census 2011 - to a UN museum in New York before he left StatsSA at the end of last month.
The donation was a gesture to mark the end of a distinguished 17-year career as our chief number-cruncher.
The name of the 60-year-old father of three has become synonymous with our post-apartheid censuses.
He is fond of telling an anecdote about how, as a boy, he chose a name for an ox his father bought in the mid-1960s. He called the new addition to the family kraal "Census" as it came a few months before his native Lesotho was to have its 1966 population count.
Growing up as the youngest of six siblings in Qibing, a village in Lesotho's Mafeteng district, he couldn't have known about the future career in demography that awaited him in South Africa. He jokes that naming the animal was probably a "premonition".
"When my father died, I realised that he had kept a record of incidents that occurred in the village. I don't know when he wrote that. My brothers and sisters left me with that. I'm a custodian of that record. At some point I think I'll anchor my being around it," he says.
The story of how he ended up practising his craft in South Africa from September 1982, a few years after graduating with a degree in statistics and economics from the National University of Lesotho, is intricately linked to his family's politics.
The Lehohlas were active in the Basotho Congress Party, which was opposed to then prime minister Leabua Jonathan's increasingly repressive regime.
NONEXISTENT DATA
The young statistician incurred the wrath of Jonathan's regime when he helped a friend and former ambassador flee Lesotho, and was forced to leave for fear of reprisals.
"I was on the wrong side of the law and I had to leave. I was intending to go to Botswana, but when I got to Mafikeng a friend saw me and said 'Stay here' and I stayed. I tried to persuade him that I intended to go to Botswana."
His decision to stay in Bophuthatswana was out of sync with his politics. "How do I stay in a homeland? Your conscience works against you, but I made a choice and I stayed," he recalls.
As a statistician-general you can’t go
into public policy to shed light. But now I can go and advise, and say: ‘Prescribe this policy because data suggests that it can be successful’
Back in Lesotho it wasn't only his teacher parents who were politically active. His brother Lesao quit teaching for a political career, and later became the leader of the Lesotho Congress for Democracy. This paved the way for him to become the deputy prime minister for a brief period after the 2002 parliamentary elections.
His eldest brother, Mahapelo, now Lesotho's retired chief justice, found himself at odds with former prime minister Pakalitha Mosisili in 2011, when the head of government tried to curtail his powers.
After landing a job as a statistician in the Bophuthatswana bantustan, one of the things that amazed Lehohla was the lack of demographical data on black people - save for information about age and sex collected by Central Statistical Services since the 1970s.
The populations of the "homelands" were excluded altogether from the CSS's population counts from the 1980 census onwards.
The paucity of the data prompted Lehohla, who had previously worked in the Lesotho's labour department, to persuade Lucas Mangope's regime to undertake a comprehensive census of the bantustan's population in 1985.
"As a student of demography and a practitioner, these numbers were very important. When I came to Bophuthatswana I found that it didn't have these numbers and that was a shock to my system.
"When you go into the public domain, you find that the most important information is about white areas and [there's] not so much about black areas. In the 1970 census, the black population data was not processed for the full details. It was only age and sex and about five variables. The rest were not processed because they said they are too many and there were computer limitations," he recalls.
NEW INSTRUMENT
In preparation for the Bophuthatswana census, he had to revamp the questionnaire to include topics that had not been covered before about the lives of black people. It included questions about housing, water, population migration and other developmental issues.
"When I approached Pretoria about our instrument, we deadlocked. They said: 'Your questionnaire is too long, we will not be able to help you process it.' I told them I was going ahead with the instrument as I'd designed it.
"If you look at that instrument I designed in 1983 for Bop, it's the same instrument that is continuing now in StatsSA - quite distinct to what the apartheid guys were constructing for their censuses. It was interesting that when we published our report in 1986, the [Development Bank of Southern Africa] were coming to see us and saying the report is very good, unlike the South African one," he says.
As South Africa moved closer to democracy in the early 1990s, Lehohla locked horns with CSS head Treurnicht du Toit over how statistics should be collated in future. Du Toit was in charge of the census in Pretoria while Lehohla was responsible for the bantustan.
And then democracy arrived and, a year into the new political order, the position of the head of the statistics agency became available.
Lehohla, Du Toit and Mark Orkin were in the running.
Orkin eventually got the job and recruited Lehohla as his head of demography at StatsSA.
Orkin also brought in a young Risenga Maluleke, from the Gazankulu bantustan, to be part of his new team.
Lehohla tells how, soon after Maluleke joined, he walked over to Orkin's office and said: "I think we've got a future statistician-general."
THUMB-SUCKING
Many years later, Maluleke was to succeed Lehohla as the statistician-general.
Statistics illuminate virtually every nook and cranny of our lives - including marriage and divorce, employment and unemployment patterns, how we name our children and the leading causes of our deaths - and the paucity of reliable data could affect both public policy choices and planning. It is difficult to imagine that civil servants could make public policy choices without relying on hard data.
Yet this happens frequently. The Sunday Times recently reported that some newly built schools in the Eastern Cape faced closure because they had too few pupils to teach, suggesting that officials ignored data when planning to build the schools.
Lehohla points out that the history of planning is littered with examples of "thumb-sucking", saying bodies such as municipalities tend to "suboptimally use the data our bean counters put at their disposal".
He cites the 2002 presidential urban and rural nodes selected for development under former president Thabo Mbeki as an example. That probably explains why the poverty these interventions were meant to eradicate is still very much a part of the South African landscape.
"[It was] a very good benevolent thought. President Mbeki's government didn't have the data then because we hadn't released the [2001] census data at the time. You look at the nodes and their choice and say: 'Oh my God, this was just a thumb-suck,'" he says.
"There's not been any follow-through around what happened in those nodes. It is a weakness in our planning systems. When you plan, you plan for the long haul and you don't forget what you thought you were trying to do."
But the problem with statistics is that they can also be a political tool. For example, some in government were unhappy in the early 2000s with the way StatsSA defined employment and unemployment in its labour force surveys because the figures did not cover the politicians in glory. But Lehohla stood his ground. There's no indication that statistics have ever been tailored to suit a government narrative.
CULTURE OF TRANSPARENCY
He reckons that the law and the people in the agency play a key role in keeping political interference at bay, thereby ensuring the integrity of the data sets it produces.
"By being in that space, I had to exercise the law to the fullest. Nobody would come and influence us negatively. When [then finance minister] Trevor Manuel was there, I would send him the consumer price index at the same time as everybody. There's no privileged info because these numbers drive markets," he says.
The new statistician-general and head of Stats SA is Risenga Maluleke, whose appointment took effect this month. Maluleke has been with Stats SA for 20 years‚ most recently as deputy director-general for statistical collections and outreach. Pali Lehohla described him as “a trusted and enduring partner“,
adding: “My relay is done. I am passing the baton to a well-tested professional and leader.”
— Taking the reins
Similarly, faulty statistics can hurt the economy. Perhaps the biggest example of that is the 2003 consumer price index debacle, when the miscalculation of consumer inflation led to unnecessarily high interest rates. The correction of the faulty data, which was spotted by an asset manager at a private firm, forced the Reserve Bank to cut interest rates that year.
"I took that as a moment to be transparent and anchor a transparency culture in the institution. I appropriated mistakes and used them as learning platforms," Lehohla says of the biggest regret of his career, an incident that almost cost him his job.
One of the things he plans to do now that he's left the civil service is to wield whatever clout he has to use numbers to influence public policy. He once irked former higher education minister Blade Nzimande when he pointed out that the percentage of black university students who failed was higher than it was in the 1980s.
"As a statistician-general you can't go into public policy to shed light. But now I can go and advise, and say: 'Prescribe this policy because data suggests that it can be successful,'" he says.
He'd also like to see academic institutions, some of which he actively partnered with during his tenure, incorporate StatsSA's statistical methods and data in their curriculums.
"I have been complaining quite a lot that people are not using the data. I am sure my successor will complain similarly. So there must be somebody who helps others to use the data," he says.
