The U.N. and Statistics | Stats + Stories Episode 86 / by Stats Stories


Stefan Schweinfest was appointed Director of the Statistics Division (UNSD/DESA) in July 2014. Under his leadership, the Division compiles and disseminates global statistical information, develops standards and norms for statistical activities including the integration of geospatial, statistical and other information, and supports countries' efforts to strengthen their national statistical and geospatial systems.

+ Full Transcript

(Background music plays)

Rosemary Pennington: The United Nations statistics division is marking its 50th anniversary this year. The division is a clearinghouse of sorts for global statistical information, while also working to develop shared norms and standards for statistical work. The work of the United Nations statistics division is the focus of this episode of Stats and Stories where we explore the statistics behind the stories and the stories behind the statistics. I'm Rosemary Pennington. Stats and Stories is a production of Miami University's departments of Statistics and Media Journalism and Film, as well as the American Statistical Association. Joining me in the studio is regular panelist John Bailer, Chair of Miami Statistics. Department of Media Journalism and Film’s Richard Campbell is away. Our guest today is Stefan Schweinfest, the director of the U.N. statistics division. He was appointed to that position in 2014 but has been with the division since 1989. Stefan, thank you so much for being here today!

Stefan Schweinfest: Well thank you very much Rosemary and John for having me!

Pennington: So how do you explain to people outside the U.N. or even sort of outside the stats world of what it is the stats division does?

Schweinfest: Well, again, thank you very much for having me, and not being afraid of the United Nations or statistics!

(Collective laughter, voices overlap)

Schweinfest: Usually when I have a dinner conversation, when I say I am a statistician, I am watching the reaction very carefully.

(Collective laughter)

John Bailer: And that ends the conversation frequently, right?

Schweinfest: Exactly, then I switch the topic. I sometimes say I work in Information Management and that sounds a lot more sexy to them. At least they give me a second chance to say what I am doing. No, I mean thanks! The U.N., if I may, just a small clarification, you got that right. With the 50th session of the Statistical Commission, but the statistics program in the United Nations goes back to the very beginning in 1947. At the beginning, we had sessions only every 2 years, so that's why we have the 50th session now but in total we already have had 72 years of a statistics program at the United Nations. So very early on when the United Nations was created, they realized that it would be a good idea to have a statistics division somewhere to gather this information from countries around the world so that we better understand our members states that make up the United Nations, so that we understand how many people there are and what their level of development is. So the initial mandate of the statistics division of the United Nations was really to gather basic information and that was originally mostly economic and demographic and social information. But in the meantime it has grown a lot and it's been following the development agenda, and now we collect really, information from all the 193 member states on economics, environment, even government arrangements and so on and so forth. So it's very interesting, what we have here in our database at the U.N. over the last 70 years from almost 200 countries.

Bailer: So how do the members use the information that you're collecting? I mean what is the value of having a statistics division that's collecting such international data?

Schweinfest: I think the value added that we provide is really the sort of comparability between data that is coming from countries around the world. So primarily, in the 1st step we collect the information from the countries. We don't add anything, so that information is already available somewhere. But I mean we're putting it all together in big table, global tables, so that you can compare between countries. But that is an important issue. Comparability is of course only ensured if we are all using the same methodology or put it in the same statistical language. And that is the 2nd big job that we have, apart from collecting data simply, is to develop methodologies and guidelines and international global norms on how a census is to be conducted or how national accounts is compiled, so that actually the data that are coming in from Uruguay or Central African Republic or Myanmar are actually meaning the same things, so that we can literally communicate from one country to the other. So then these data can also be used in international discussions, for instance, for our General Assembly, when they have a development agenda which is the situation that we have right now, with sustainable development agenda, and they want to know how much progress have we made, and do we need to pay attention to either particular groups of countries that are seemingly falling behind or to particular sectors of the development spectrum where we seem to have more difficulties than in others. So in that sense our data are really supporting the policy dialogue that is taking place at the United Nations and elsewhere. Of course, we are also using the data not only for the U.N. but for any citizen, because we have them all on our website. So anybody who's interested can go there.

Bailer: OK. Let me follow up and ask you to dive into a particular example. So this is sort of a general description of what you're doing. Can you give an example of a particular variable that you're measuring in different countries? And it would be really interesting to hear about one that's difficult to define and difficult to measure in some consistent way.

Schweinfest: Let me just start with an easy one.

Bailer: OK! Please do!

(Collective laughter)

Schweinfest: The National Account System was developed and now G.D.P. is universally used and it means the same thing. I got a phone call, just also to make the point that our data are not only for the politicians but it’s also for the private sector. And so I got from a company, a US company that said that they are considering investing in Southern Africa and they wanted to have data, all the economic data that we had for 10 African countries where they considered perhaps opening a plant. So that was an interesting case for our services, also the private sector with our information because they wanted to have data on growth, on development, on trade, on the size of certain subsectors of the economy, how big is the oil sector, how big is the agriculture sector, and so on and so forth. So all of that information, we could find, and pretty comprehensively because that's the information we have been collecting for over 50 years now. New indicators that are difficult to…more difficult is the whole spectrum of environmental indicators, I mean how do we measure air quality or biodiversity, that has been developed over the last let’s say 10 to 20 years. We are now at a point of great convergence. And the newest indicators are more the structural or institutional or governments’ indicators, so for instance if we want to measure, does a country have free press, if somebody asks me that and says how many free newspapers do they have or publications. Well that's difficult to define because how do you really measure that exactly. So these are the topics that are being increasingly discussed, but we are still in the process of trying to develop language so that it’s really meaningful to compare the information that we are getting from different countries.

Pennington: I was going to ask you how difficult it is to get, you know, 200 countries to all sort of ride together on agreement, on sort of the measures, like how do you marshal this sort of interaction amongst countries to get them to agree on something like this?

Schweinfest: It is very important that we all agree on that because I mean, the United Nations doesn't have any power, I mean to tell any country around the world to do anything. I mean, you know the only power that we have at the United Nations is what I call the convening power, to invite people around the table and say let's all sit together, let’s tell the stories of how you have tried to approach measuring that, what good and bad experiences have you made with that and what could be a good way of making a recommendation of how this could be measured globally. So we do have usually what we call expert consultations where we really invite representatives from countries around the world and say, what have you done to measure gender equality. And then we make sort of an inventory of current practices and then we try to move on to good practices or best practices and usually after a couple of years or so we're talking long processes. I mean these are not things that happen overnight. We usually get to an agreement and a recommendation and most countries are quite willing to do that because it's in their interest also not only to develop their own measures where they can communicate within the country because I mean they also want to broadcast beyond their borders what is the situation in their country because there’s international development aids, there are international agreements that they are part of, and there’s the private sector that also tries to understand the citizens around the world. And increasingly there is also networks of civil society organizations that also want to understand where, in which countries there is a particular need, where they could perhaps be particularly active, and so on and so forth. It's been quite exciting, the diversification of uses in statistics over my lifetime.

Bailer: You know I like the slogan that's associated with your group, this idea of better data, better lives and I was wondering if you could talk about where that came from and if you could give examples of where you think that better data has led to better lives, for the members.

Schweinfest: I honestly don't remember now, I mean somebody…

(Collective laughter)

Schweinfest: who had that brilliant idea of… you know how this goes. You sit a group of people together and everybody is throwing something on the table and you choose one and then suddenly you have the moment where they think that sounds really good. And we've been using it for a couple of years now but I mean mostly the data that we collect is to support development, I mean health programs and education programs and so on and so forth. So I mean if you have information on, for instance, the number of schools in a particular country and beyond that, one topic that has been becoming very relevant is the connection between statistical and geospatial information. So if in addition to having information on the total number of schools you even have a map of how these schools are distributed in the country because it makes a big difference whether all of your schools are just in the rich sector of the capital or they are really distributed all over the country including the rural areas. Then you can get a very good picture of where you need to invest. I mean the government or the international community that may be actively supporting the country, where they have to put the effort. I mean, to invest in schools, in infrastructure or in the health sector, in the campaigns, health campaigns,(inaudible) and so on and so forth. There are numerous examples I think where having better information and being able to target especially health and education interventions have produced tangible results, really for rural communities in developing countries,

(Background music plays)

Pennington: You're listening to stats and stories and today we're talking with Stefan Schweinfest, director of the United Nations' statistics division. You were talking a lot about the development work that you gather data for and I know that a lot of times journalists will cover stories about sort of development projects the U.N. is involved with. What do you think are under covered stories, sort of sitting in the data that the U.N. is collecting that maybe reporters ought to be doing more digging into?

Schweinfest: Well you're putting your finger on something that I sometimes feel a little bit frustrated about and I'm also pointing to ourselves as the culprits I mean we are usually as statisticians, very concerned to get the numbers right and we polish our numbers and we are very proud when they are coming out. But then I always say, this is not the end of the story. I mean we need to use those numbers. A good number is only a used number. As you challenged me earlier to give you examples, how numbers really made the lives better by being used in a policy decision. So I think journalists play a really important role because they help us with the communication, they are intermediaries and so they help us also to dig out from their perspective what they are interested in understanding of users, what are the interesting elements in all of those numbers that we are putting in our database. So I think I regret sometimes that we do not have more interaction with journalists that help us actually to bring out those stories. Now with the sustainable development agenda, the universal focus has been on certain development goals that have now been agreed upon among the member states. So here I think we are now developing almost a culture of communication, which I am very happy about. So that we are not only just churning out our numbers and then go home and feel good about that. We also feel responsible of bringing those numbers to the people, helping together with the users and the journalists to reflect what is the story those numbers tell. I mean, what does it mean that those numbers went up and down again etc.? And what can be done about it. So I think journalists are very important partners for us but of course sometimes we are living in a complicated world, lots of numbers are flying around these days. And one important topic that is close to our heart is of course quality.

Pennington: Yeah.

Schweinfest: Data quality is something that is very important because two numbers that are out there, they clearly don't necessarily mean the same thing.

Pennington: Right.

Schweinfest: Because one number may be a very rough estimate that somebody came up with in their head, and another number is the result of a very thorough statistical process where a survey was conducted, and where we know that this number is really authoritatively reflecting the situation out there.

Pennington: So what advice would you give a journalist who is sort of trying to dig through all this data from various sources as they’re trying to write a story that is responsible? How should they go about sort of making sure the data they're using is quality data?

Schweinfest: Well I mean one of the fundamentals is always of course is sourcing the information. I mean I’m always a little bit annoyed when I see numbers anywhere and I'm not seeing where that number is taken from, so I mean that is sort of number one rule of the game. Then the second thing I would say I mean is I think we need to all get into more of a habit of a dialogue with the data owner or data producer because…the data custodian is a good word…because we know how solid that number is and I also know that not all numbers that we have in our U.N. data base are equally solid. Not all countries are equal, not all subject matters are equally developed. Sometimes a 5 really means something between 4.9 and 5.1 and on another page a 5 may mean it's more between a 4 and 6 and so it's the people who have worked with the data, they know, they can explain a little bit about the confidence interval around those numbers and that's kind of also I think that needs to be communicated also to the users and I have great confidence that people are quite capable to understand, digest nuances about the numbers. They do that all the time in their private life. I mean if we all are told by the weather service that it is going to be 50 degrees out there, we will not sue them if it’s 49. We understand that that is an approximation and it sort of gives you a guidance of whether you need to grab a small or a big sweater. With data it's often the same. Some are very precise and others are more approximations, and I have great confidence that people can understand the difference,

Bailer: Just a follow up on that, I mean, how do you communicate some of the uncertainty? I mean you're saying that some variables that you are reporting have differing degrees of precision and some of the data from certain countries have different levels…you have different levels of confidence in the accuracy and precision of those values. So how do you make sure that that message is put forward?

Schweinfest: Oh, well we have a very annoying way of doing that!

(Collective laughter)

Schweinfest: And they are called technical notes, which I of course know, nobody reads. But I mean that is basically the answer to your question in a way. What also happens is when a country gives us information, we are getting government information. I have no right as a UN official to change that number. Also, who am I in New York, even if I have a strong feeling that number is not accurate, to impose my vision of the world. But what I can do is not publish the number. I am under no obligation to pass that information on to the rest of the international community in a global table if we are not satisfied with the quality and we go back, and a key word there is metadata. We ask, how was that compiled? What was the sample size? Was the data only collected in the capital or was it also collected in the rural areas, and so on and so forth. What was the time frame? What was the coverage? And so what we are asking are all kinds of annoying questions and if we do not get the answers, we can decide that number is not yet mature for international publication or we would have to put a serious footnote where we say, there were some questions that were not answered, for the educated reader, a bit of a warning sign for the confidence interval of that particular number is probably quite a bit bigger than all other countries.

Bailer: I'd like to follow up. You mentioned sustainable development a couple of times. Could you just describe generally what the sustainable development goals are and what the U.N. is doing to work towards these objectives?

Schweinfest: Yeah of course that is another development area. I think our big common topic right now and I think in a way I am relatively proud also of the United Nations that we have been able to reach that degree of agreement because I mean, everybody understands a little bit different under the big term development. But I mean in a rather extensive negotiation process over 2 years from 2013 to 2015, the international community here has agreed on a development agenda framework in particular 17 particular goals that cover the areas of economic development, social development, environmental development. And they have codified the goals that we are trying to reach as a global community by 2030. For instance universal primary education or gender equality and then it's broken down to more specific targets, and that is now as an action agenda, underpinned by over 230 indicators that we are now trying to measure to help countries in the global community to actually understand whether we're making progress towards these goals and targets.

Bailer: So which goals has there been the most progress on, since this has been framed?

Schweinfest: Oh I mean, one should say that the sustainable development goals build on what was there before, the millennium development goals. That was the development framework agenda that had been agreed in 2000 and ran until 2015. So I mean we didn’t quite start from scratch. But that development agenda was much more reduced, there were only 8 goals. We have now moved to 17 goals so it's a much more comprehensive agenda which also has more visionary targets like a just and equitable societies, which are noble visions, but it's not so easy to translate that into real concepts of measurable elements. But I would say the traditional area, where we have been working, reducing extreme poverty, eliminating certain diseases through extensive campaigns like polio or malaria or reducing incidents of malaria and so there I mean enormous progress has been made over the last 2 decades, and that can be very specifically measured and reflected. Also access to education everywhere around the world.

Pennington: So Stephan we certainly live in a moment when it seems like a statistical information is in doubt. Information from sort of official sources is often doubted. How has this sort of growing skepticism, changed the way the UN is attempting to communicate about their data, it's collecting and that it has available? Has it changed the way you sort of, frame the work you do?

Schweinfest: Oh yes it has. I think, I mean me personally, I learned a lot from my father who was a marketing director and he has always reminded me that it's not only the product and its quality, it is also the dissemination, distribution and the connection with the user, that is important for you to successfully place your product. And I think that has taught me personally a lot, because I think previously, we as statisticians, we were just satisfied, it was almost like a more scientific approach to produce a number and we didn't really care so much, we didn't consider ourselves the public relations managers for our own numbers and I think that has changed a lot. I think we feel a lot more responsible now to go out and actually also explain and defend our numbers, not just put them in a database or in an international year book or something like that and then it just stands there and perhaps a little bit naively assume that we have done good and we have done our homework and it will automatically be rewarded. I think we are a lot more in the business of communicating our information, and sometimes also criticizing other information because I mean it is a competitive place. If there is, I mean we have something that is called the fundamental principles of official statistics, which is a kind of I sometimes call them the 10 commandments of good data or good information. I mean this is something that's been around for 25 years. So this is nothing new, and it talks about data, how to reproduce by strictly scientific methods and so on so forth. And one of the principles also says that we have the right as statisticians, as statistical producers and perhaps even the obligation to comment on misuse of information so if somebody takes our data and says this means this, and this, and that, and it doesn't, then we do have, really not only the right, actually almost the obligation to say no. That’s an erroneous interpretation, or challenge the source of another data item in the political debate. I think that’s an important contribution that we have to make, also through an educated conversation in society.

Bailer: Stefan, we can hear a clear passion for what you do and the work that the UN statistical division does. What advice would you have for people that want to get involved in working in this world, you know, for students or for others that would love to be involved in these types of official statistics, in particular the international applications of official statistics?

Schweinfest: No I mean you're right about that. I warned you before, if you get me talking about these things you have to find the stop or eject button!

(Collective laughter)

Schweinfest: I have walked into this building 30 almost 30 years. This summer I will have my 30th anniversary here and I've really learnt a lot all along, and I have always been passionate about this. I of course, like international affairs. My father always accused me of professional tourism! I love languages, I started in London and Paris as a German, so I mean I was already at that level, very interested in moving, in getting out of my little village. In the course of my 30 years I've just sort of…being a statistician, of course I keep track. I’ve had the opportunity to visit 112 countries. And I consider that…really I don't want to brag about it, it's really an enormous privilege. I have met so many wonderful people who are dedicated to producing numbers under the most difficult circumstances, where sometimes there was civil strife. When we think of a census, it is sort of a cute exercise. We send somebody an email or in the worse of all cases, we send somebody to knock on the door with a questionnaire and collect information. But there are many, many millions of people that live in very difficult circumstances, remote areas or unsafe zones of the world and they are all people who work as statisticians around the world, and they try to get that information, they try their best to keep the scientific standards up and collect that information as best as they can. And that's where I draw my real inspiration from. So I mean it's been really a privilege. I entered the statistics division through what was then called the National Competitive exam. So if you have a Master’s Degree in the area of Statistics or Economics or Social Science or related, you can take part in an international competitive exam. And then I was lucky enough to be offered a junior professional position here in the Statistical Office in New York and yes, this is where I have been all my life. And I am still annoying the people here!

(Collective laughter, background music plays)

Pennington: Well Stefan, thank you so much for being here. That's all the time we have for this episode. It was a real pleasure to talk with you today!

Schweinfest: No, thank you very much for letting me talk about what I really enjoy to do every day of my life.

Bailer: Thanks Stefan.

Pennington: Stats and stories is a partnership between Miami University’s departments of Statistics and Media Journalism and Film as well as the American Statistical Association. You can follow us on Twitter, Apple's podcast or other places where you can find podcasts. If you'd like to share your thoughts on the program, send your e-mail to Or check us out at and be sure to listen for future editions of Stats and Stories, where we discuss the statistics behind the stories and the stories behind the statistics.