Stefan Schweinfest, Director of the United Nations Statistics Division: Exclusive Interview

Stefan Schweinfest, director of the United Nations Statistics Division

In an era of fake news and populist politicians Stefan Schweinfest, the head of the UN’s statistics operation, is well aware of the threat to evidence-based policy making. But this is also an age of opportunity for statisticians, he tells Matt Ross – with growing collaboration, new data sources, and the crucial task of monitoring progress against the UN’s Sustainable Development Goals

“Some people, either for irrational reasons or for malicious political reasons, want to discard evidence,” says Stefan Schweinfest. “We need to stand up and say: ‘You do that at your peril!’ And I notice a new professional solidarity between all of us who work in that space of evidence and analysis, because we have these headwinds. We need to explain better what we produce – which in our case is data and information – and help people understand the very complex reality.”

Four years after his appointment as Director of the UN’s Statistics Division, Schweinfest is plainly concerned that data is losing its power in the battle of ideas. Statisticians, he argues, must do more to explain and “sell” their work – injecting hard evidence back into the public debate.

“We run the only production process where we leave our product at the end of the assembly belt, and don’t open the factory gate, carry it out into the sun and advertise it,” he says ruefully. “If you produce a car, you have a marketing process where you explain to people how to use the car – what it can do and what it can’t. Perhaps previously we statisticians thought that because we’d invested so much effort in the product, its value would be self-evident; but nothing in this world is self-evident. If you believe in what you’re doing, you have to make the effort to go out and explain it.”

Making evidence effective

Schweinfest is – as this interview demonstrates – doing his bit to raise understanding and awareness of the importance and power of good data. And to some extent, he believes, those who disregard strong evidence will be found out: “They’ll come up with their own numbers, and they’ll self-destruct; they won’t stand the test of time.” But he argues that statisticians alone can’t roll back the tide; and that wider changes are required in the ways that both data professionals and governments manage and present information.

For example, he says, statisticians have traditionally been very cautious in releasing interim data. “But an approximation early on can be much more helpful than exact number later down the road, when it’s too late,” he says. People understand that early GDP estimates may be revised as more information comes in, he adds, yet “we haven’t been smart enough to use that mechanism in other areas. We need to be much more flexible in selling what we have, at the right time and with the right footnotes.”

Another “very important element of generating trust”, says Schweinfest, is government’s establishment of “autonomous national statistical systems that are robust to changes of government.” It must, he argues, be clear that politicians can’t influence the data published by departments and statistics offices: “Everybody needs to know there’s something we can rely on, irrespective of all the funny things that may be going on in the political space.”

Widening the net

But whilst official government stats must be absolutely robust, Schweinfest also believes that statisticians must embrace the fast-growing wave of information produced outside governments – making use of the data produced by businesses, charities and individuals to augment public sector records and shape statistics professionals’ advice to policy makers.

“Statisticians used to be a bit stand-offish, saying: ‘If it wasn’t produced by me, it can’t be any good’,” he argues. “But we have a role in helping to make sense of the information that’s all around us, evaluating it and making sure the right information gets to the top. Statistical offices then become advisers to ministers on whether that information is solid enough; we become service providers as well as goods providers.”

After all, he adds, the private sector has “better technology, better data miners and data scientists” than most of the public sector. Civil society and academia also produce lots of strong information; and these days, citizens themselves produce a wealth of digital data. “So building that information pyramid, running from the very local level to your global databases, is one of the exciting tasks that we have right now,” he says.

Sustainable statistics

“…I don’t want to be Mr Indicator any more; I want to become known as Mr Capacity.”

Schweinfest’s office has, for example, begun using some academic data to feed into the indicators supporting the Sustainable Development Goals (SDGs): it’s his team’s job to generate the metrics against which progress will be measured. And this, above all, is the issue that dominates his inbox: with each of the 17 goals incorporating a series of specific targets, along with milestones running out to 2030, producing robust statistics on a global scale is a truly massive task.

Whilst the goals and targets were set by the General Assembly in 2015, the indicators have followed along behind. “My argument was: let’s invest and get this right, even if we are lagging a bit behind, rather than doing some ad hoc fixes now which we may regret five or seven years down the road,” he recalls. The risk, he explains, was that ill-founded indicators would lead UN member states to focus their efforts in the wrong directions – with the result that the General Assembly’s true goals wouldn’t be realised.

Whilst the right information-gathering systems are put in place, Schweinfest’s statisticians are using “proxy indicators” to help member states get moving on the targets. And when the final indicators are up and running, he’s confident that they’ll represent a major improvement on the metrics used to track progress against the Millennium Development Goals (MDGs) – the SDGs’ predecessor.

Getting the right numbers right

In part, this is due to serendipitous timing. “When the MDGs were formulated in 2000, it was very unfortunate that the 2000 census round had just been conducted,” he recalls. “We realised a year later that we may have some modified information needs.” So this time, his team is working with national census teams to ensure that these huge data-gathering exercises produce the stats required for his SDG indicators. “This is exactly the moment when many countries are designing their censuses, and we have spent a year or two figuring out our information needs,” he comments. “And this will be the first census round that’s geo-coded – so afterwards we won’t only have the total numbers, but also their geographical distribution.”

The result of all this work, says Schweinfest, will be “a data cube: it involves 193 countries and 240 indicators, then we build that up floor by floor over the next 15 years. But of course at the moment that cube looks more like a Swiss cheese: it has enormous holes in it!”

Plugging those holes will involve looking beyond government sources – but there is also much that governments can do to improve their own use of data, says Schweinfest. “It’s very frustrating that we sometimes find that in a country there’s more information than people think – it’s just hidden in somebody’s department,” he comments. “Sometimes, government officials sit on their data for the wrong reasons. How do we liberate all that information?”

Data party: all welcome

Building the right legal frameworks can help, he says; his office is collecting examples of robust data-sharing agreements, with the goal of helping governments to “make more use of their statistical information and not get lost in turf wars.” And it’s important to create effective central data structures: he praises Mexico, which he says is “well equipped because cartography and statistics are in the same institution” – enabling the country to link geometric information with other forms of data.

This bringing together of different databases and types of data is, he argues, crucial to improving policy making and service delivery. “If you improve education, you could also have better health and productivity,” he points out. “Any intervention will have ripple effects, some of them good and some bad, and we need information systems that are so linked together that you can evaluate their ultimate impacts.” Combining databases does bring the risk of ‘de-anonymising’ data, enabling people to identify particular individuals – but these barriers are not insurmountable. “Australia has married statistical and geospatial information responsibly,” he notes – protecting people’s anonymity whilst providing public servants with a clearer picture of the population’s needs.

Nonetheless, he continues, governments must tread carefully here. “Data is power,” he says. “With great power comes great responsibility – and, of course, also greater risk.” Public servants around the world must develop a “data culture: the whole package of dealing with data responsibly and having legal frameworks and social conventions in place.”

“…nothing in this world is self-evident. If you believe in what you’re doing, you have to make the effort to go out and explain it.”

Let’s get together

These days, all of these agendas come together at the UN World Data Forum: a global event first held in Cape Town in 2017. Involving governments, businesses and civil society, the first Forum produced a Global Action Plan which, Schweinfest recalls, focused on “strengthening national statistical systems; working together across government boundaries; raising funding for data and statistics; and educating users to use numbers wisely.”

Though these goals were quite high-level, he adds, the Plan has “created a platform for dialogue” and lead responsibilities for delivery have been allocated. The second Forum will be held in Dubai in October, he explains: “This will be the first time that the same set of people get together, and we can take stock on how it’s working.”

Meanwhile, Schweinfest is shifting his own focus. He’s set the indicators underpinning the SDGs; now the priority is to help governments develop their data gathering and management operations, populating those metrics with strong, reliable data against which the world can measure its progress with confidence.

“The formulation phase is over; the indicators have been defined,” he says. “But there are still many countries which don’t have the financial or technical resources; and that defines where we focus our energy. Now the name of the game is giving countries the capacity to compile that information – because if we don’t do that, we have a real problem.”

So the UN Statistics Division is moving into a new phase; and so is its director. “When I walked through the streets of New York, at one point they used to call me Mr Indicator,” he concludes. “And now I’ve officially declared that I’m done with that. I don’t want to be Mr Indicator any more; I want to become known as Mr Capacity.”

Related Posts

• 

  UN introduces feminist climate justice framework for policymaking
  UN introduces feminist climate justice framework for policymaking
• 

  Number of women and girls in conflict-affected countries doubles, UN report finds
  Number of women and girls in conflict-affected countries doubles, UN report finds
• 

  UN calls on countries to include women’s reproductive health in climate plans
  UN calls on countries to include women’s reproductive health in climate plans
• 

  UN members see need for ramped-up progress on gender equality
  UN members see need for ramped-up progress on gender equality