Episode 2: Strengthening national data ecosystems to leave no one behind

In the first mini-series of Good with data we explore one of the most important issues in global development today, the Leave No One Behind Agenda; what it means, why it matters, and how we can make it a reality by improving data and making best use of existing data and evidence. 

In this episode we discuss national data ecosystems, and why they are key to fostering a strong culture of data use to improve development policies and programmes that leave no one behind. 

Our guests are: 

• Elizabeth Birabwa Aliro, Programmes Manager at the Economic Policy Research Centre in Kampala, currently focused on strengthening Evidence Informed Decision Making in policy processes in Uganda; 
• Papa Seck, Chief of Research and Data at UN Women, where he leads the organisation’s research and statistical work, including the Women Count global gender data programme.

For more on this subject, read our recent discussion paper on the role of national data systems and data to leave no one behind. An accompanying discussion paper looks specifically at the role of donor investment in strengthening national data ecosystems, and how they can better act on their data commitments. During the episode, we asked our panellists to share their recommendations for listeners to explore issues relating to national data ecosystems further:

• Elizabeth recommended two recent reports from the Economic Policy Research Centre: one on evidence diagnostic exercises in Uganda and Pakistan undertaken as part of the Strengthening Evidence Use for Development Impact programme; and another on the political economy of policymaking and the evidence ecosystem in the Humanitarian, Gender, and Family Planning sectors in Uganda. 
• Papa recommended visiting data.women.org to view UN Women’s relevant resources, and encouraged all of us to consider how decisions based on data affect our lives, and to get involved to make sure that the data being used to make those decisions is good. 

Good with data is a production of Development Initiatives, a global organisation harnessing the power of data and evidence to end poverty, reduce inequality and increase resilience. 

To stay up to date with our work, follow us on Twitter or Linkedin, visit our website, and register for email updates.

We value your feedback! If you have comments or ideas for the show please contact us. If you enjoyed this episode, please subscribe and leave us a 5 star review wherever you listen. 

Transcript

Deborah Hardoon: 0:03

Welcome to Good with data development initiatives. I'm Debra Hardoon. And in this three part miniseries, I explore one of the most important issues in global development today, the leave no one behind agenda, what it means, why it matters, and how we can make it a reality by improving data and making best use of existing data and evidence. In this episode, I'm speaking with Papa Seck. Papa is the Chief of the Research and Data section at UN Women, where he's responsible for the development and planning of research and the use of this research and evidence on gender and women's empowerment. We also have Elizabeth Birabwa Aliro. Elizabeth is the Programme Manager at the Economic Policy Research Centre in Kampala, Uganda. Her current focus is on strengthening evidence informed decision making in policy processes in Uganda. Today, I'm talking to Papa and Elizabeth about the infrastructure, the finances the systems and politics that's necessary for data on people and their outcomes to be of good quality, accessible and useful in the pursuit to leave no one behind. In the other episodes in this series, guests from across UN institutions and civil society have shared their perspectives on the leave no one behind agenda. In one episode, we discuss how data on risk can be helpful to help us think about who may fall behind as a result of shocks and crises. And in another, we cover her data and evidence on inequality can enable us to better identify and tackle the root causes of poverty. But for this episode, we're heading to the back office, we're not talking about what data tells us about how people are doing and who is left behind. But rather, we're looking at what needs to be in place in order for good quality data to be collected, to be produced and to be shared and critically to be used. So before we get into discussing some of those foundational requirements, I want to ask you both Papa and Elizabeth, what data investments, the laws, the systems, the processes, what are they seeking to achieve? In other words, what does useful data look like to you, particularly when it comes to supporting efforts to leave no one behind? Maybe to you first, Elizabeth.

Elizabeth Birabwa Aliro: 2:23

Thanks, thank you. What comes to mind when I think about what good data would mean, is the fact that it has to be generated and shared. And it has to be nationally representative. And by focusing on nationally representative, is to ensure that it covers also those who are normally marginalised populations that tend to be left out, sometimes when big surveys are being conducted. So for me, it is data that is generated and shared in a nationally representative manner. It is also data that is multi-themed, that is looking at different aspects of an issue. If it is poverty, for example, then it should be looking at all the multi facets of poverty. And this data is publicly available to ensure wide use by various stakeholders. But most important for me, for data to be valuable, it has to be then used by the critical people who make policies and decisions. And here are mainly looking at big data or evidence being very relevant to the politicians and policymakers for purposes of planning, budgeting, programming, and monitoring and evaluation. And if the policymakers then find this data relevant and use it, then we will have service delivery that is best or designed based upon good data and good evidence. And with this kind of service delivery, then everyone will be incorporated. So in such a way, no one will be left out within this equation.

Deborah Hardoon: 4:29

Thank you, Elizabeth. I think that's really powerful that you're defining good quality and useful data actually in terms of the outcome, so whether or not you know, people are better off and, and services are better informed by evidence. Papa how about from your perspective? What does useful and good evidence look like to you?

Papa Seck: 4:49

Yeah, no, thank you very much, Deborah, and thanks for having me as well. So I will actually start where Elizabeth left off from which is you know, that data, that is good data to me starts with data that is used to change people's lives, right. So there is no use producing very good data if that data is not used. Again, as I usually say, useful data is data that are used. So, to me, I think that's basically one foundational pillar of the work that we do, understanding that the work that we do really has to contribute to changing better people's lives, better lives. But I will say that also, I think, in my area of work, that one of the fundamental things that we see is that a lot of the data that currently exists, you know, I think it's good enough to tell us more about, you know, what needs to be done and etc, give us solutions. But I think unfortunately, also the data can be quite biassed. And we call it sexist data in my area of work. Essentially, I mean, when we say when I say, said earlier that data has to be used, it has to be good data, it has to be data that reflects gender inequalities, and really doesn't, is not based on norms and gender stereotypes in its production. So making sure that data really reflects the key principles of gender statistics is fundamental in what we do. Because, again, bad data is actually can be more dangerous than, than any data at all. And that's the work that we tried to do around, around the world to make sure that surveys really reflect adequate gender equality issues and principles so that they can be considered good data for use. But, you know, again, I think the final point is really around, you know, making sure that the data that is produced is owned, at national level. To me, that's also one fundamental aspect of the work that we do, understanding that national statistical offices play an important role. But there is also out there, I think, a whole ecosystem of stakeholders that need to be brought in, so that the decisions on what data are produced and what data are used are made collectively, and not by a select few. So that really issue of you know, ownership, participation is really important in order to make sure that that data that we use is really reflects essentially the people's lives,

Elizabeth Birabwa Aliro: 7:31

maybe to add, just to build on web per sec. As for the issue of the citizenry, that normally yes, we do have survey data, we do have all these forms of data. But the citizens' feedback is also very critical in terms of ensuring that we design programmes and interventions that meet their needs. And often, in this part of the world, we rarely do invite citizenry in our data collection processing. We do engage them as respondents, but often not involved in the design of some of the survey instruments, and also in the other data processes such then when we do produce the data out of the feedback that sometimes they've given us, you find often that they too cannot use this data, they are overwhelmed by the kind of data that we've collected by them. And therefore they fail to use it to shape the way they want to devote their community. So I wanted to emphasise also that need for citizens' feedback, and how to also be able to involve them in the processes of data gathering and data processing.

Deborah Hardoon: 8:51

Elizabeth that really does lead me to my next question, actually, which was all about how we think about good data from a leave no one behind perspective. So, often we think when we're talking about leave no one behind that we're really interested in whether or not data can be disaggregated, by gender being an obvious inequality, but also other locally relevant inequalities of ethnicity, or religion or sexualities and so on. So being able to cut the data to understand particular individuals or groups that may be left behind. But another aspect of data for leave no one behind compels us to think beyond the statistics themselves, and think about power and inclusion. And that this inclusion is important in the data sphere too in the data processes with people having a voice and a role to play throughout the data collection, management and analysis. So I think this very much builds on, on the point you're making about citizens and I was wondering where you think we are in the data ecosystem world that we're working in, in terms of really meaningfully including people in these data processes?

Elizabeth Birabwa Aliro: 10:02

I would say attempta are being made. But I don't think we are yet there, especially in meaningfully involving people in the different, the citizenry and other stakeholders other than non-state actors. In Uganda, you find much of the data processes are so much slanted the data gathering processes, the data management processes are so much slanted towards the state. And you find that there is low engagement, involvement of the non-state actors. You find like right now, the civil society organisations do generate a lot of data, but you will find this data is not absorbed within the national statistical system. And the argument that is always fronted is that either the scope is too narrow, or the the designs of the data collection instruments were not well trimmed. And therefore, all the issues are very [inaudible] and micro in nature and therefore you find out, and yet the NGOs are the ones that normally deal with the issues, where are they vulnerable, where the gender dimensions, the gender issues all come to play. So, if we leave out, we have no framework that guides how the civil society organisations, kind of data collection activities can input into the mainstream national statistical system, then we are in a way, leaving out, we are not involving everyone.

Papa Seck: 11:44

Yeah, I think, on that question were really, again, we've made some headway. I remember, you know, the conversation that we had before the, during the MDGs, and the conversations that we're having now, and it's night and day, so really, you know, the fact that I think there is a wide recognition that, you know, the data ecosystems are much larger than, you know, just the national statistical offices themselves and the work that they do. And that, you know, there are lots of stakeholders that need to be included, for instance, I mean, in, in Uganda, in fact, I think, you know, the work that we've done there with the Uganda Bureau of Statistics, but others is really create a national coordination mechanism that brings together, you know, various stakeholders in the data ecosystem itself, national statistical system, so civil society, academia, and so on, line ministries, etc. And those coordination mechanisms really do help. In terms of defining a broader ecosystem, we've done that in Kenya. And what we've seen also is that once you have those structures in place, it really helps in terms of both improving the quality of what is produced, but also improving the use. But also, you know, part of the work that we've been doing is really on citizen-generated data, in fact, in Uganda was our first pilot country in that regard. And, you know, the guidelines that we are coming out with will basically address essentially, you know, what needs to be done. And it's really that exchange between stakeholders, but also international statistical offices, which play an important role. We see them as data stewards now, really, I think, bringing together a whole ecosystem of stakeholders, and I think that really will help us to advance so I'm really optimistic in the sense that I think we are in a much better place now than the conversations that we had just, you know, five years ago. But, you know, again, I think there's still much to be done, because issues of representation loom large, for instance. And here, I'll just give you an example. When it comes to gender, we see, for instance that, you know, if you look at national statistical systems, national statistical offices, chief statisticians are by and large, all men, the representation is all male. So, you know, unless we start diversifying how our governance, data governance systems work in terms of their representativeness and etc, we will not achieve, I think, the results that we are hoping to achieve. So I think, you know, again, one area where we really need to make some progress is the issue of representation to make sure that it is diverse enough, it is gender diverse, that also includes other groups that need to be at the table.

Deborah Hardoon: 14:43

Things that really does emphasise the importance of looking across the data system, not just at the data itself or even the collection, but at the whole infrastructure and who indeed has the seat at the table in those governance positions as well. Thinking about other areas of improvement, Elizabeth, to strengthen evidence informed decision making in Uganda as per your role at EPRC, you of course need to rely on the quantity and quality of evidence being generated in Uganda. So can I ask where have you identified weaknesses or gaps that have limited the extent to which you can rely on Ugandan data to inform decisions in the right

Elizabeth Birabwa Aliro: 15:29

in Uganda around the use of way? administrative data. This is data that is mainly collected at the local government level, by the local governments. And this is data where they are either monitoring indicators that have been set at the national level, or indicators set for their level in terms of service data monitoring, service delivery. And the challenge here is normally the issue of always wanting to paint a rosy picture of wanting to say we are succeeding, we are implementing, we are seeing results in implementation of this particular indicator, let's say access to water, then that kind of slanted reporting around these indicators at that level in terms of the admin, in terms of the administrative data, so you find there issues of accuracy and reliability. And this kind of data is not being so much used, other than for purposes of producing annual reports reporting for government. But other than that you don't see yet if it's well coordinated, well managed capacity is built, and then the issues of accuracy handled, then this is a kind of data set, that could be a data source that could be used to inform policymakers better than it's currently being used. And obviously, there is this other issue of data that is not generated internally, if I would codename it that, data that is generated by external sources, external funding, or external consultants, there is that feeling that this is not reflecting the true picture on the ground. Among this they policymakers, decision makers. So these are areas where we need to work towards improving the perception that's a perception issue. And this is based on a recent diagnostic that we conducted on evidence use in Uganda, and we were looking at factors that hinder or facilitate uptake of evidence, which is likely related to data. And we found out that that was a key perception; that policy actors, especially within the state, held that if the data is not generated, internally by Ugandan consultants, or by Ugandan led teams, then their capacity to absorb is low. Obviously, there is the issue now that is coming up the issue of managing big data and transmit and transactional data. Given that we are having a state, this is not something that has been happening in Uganda, we've not been having a national identification card. And then the process was started to ensure that every Ugandan is registered and issued with a national identification card and number. And then you find that this personal data is being collected by, your data is being collected by different entities, both state and private sector. But right now, what we are seeing is that we don't have a harmonious way of doing it and there is no central place where this kind of data is, is kept and can be accessed and managed. You find that each entity, each agency, Uganda Revenue Authority is keeping its own data, you find that the national identification authority is keeping its own data. So and then you find like the mobile, the telecoms that also have their own data set. So what we are saying is, how can we have a harmonious way of how we can bring all this data together in a manner that can facilitate easy access, easy management, but at the same time, security around this personal data?

Deborah Hardoon: 19:43

Elizabeth that's so interesting. In DI we often talk about the virtuous circle between high quality data and good data use with the idea that the more that data is used, the more that creates a demand for data to be of better quality, the more therefore that that data gets used. But I think what you're saying there is, is that's not enough, actually. Because sometimes the data no matter how high quality it is, is just not being produced by the right people in inverted commas. Or the data is just not telling the right story or not the story that people want to hear and that, and that, therefore, its use is really limited no matter how high quality it is. So I think that's a really a really powerful challenge, I think, to our kind of idealistic kind of picture of how to create this virtuous circle of data quality and use. I'd like to kind of turn the conversation a little bit now to be a bit more propositional and positive and solutions orientated. I mean, you've both mentioned kind of things that have come a long way, whether it's in meaningful participation and inclusion or about, you know, improving the data that's coming from non state actors, for example. But if I was to kind of push you on that, and maybe Papa to you first, if, if you could wave a magic wand to strengthen data ecosystems in a way that gets us to the kind of data utopia that we were talking about in the beginning, where would you put that wand to best use?

Papa Seck: 21:17

You wouldn't have to be multiple wands right? So one, you know, I've spoken about earlier about essentially, these national coordination mechanisms on gender statistics that we've really supported countries to establish. And I've seen them working really magic in terms of bringing stakeholders together, deciding together on what is produced, what is analysed, what questions need to be answered. And what I've seen from there is that basically, from that, essentially, data use just becomes natural, right? So because people actually do find the data to be relevant to their, to the questions that they have to be answering those questions. So using it becomes becomes normal. So I would really place a key emphasis on that. But for that to happen, essentially, you need the investments, right? So investments cannot just be in data collection, investments have to be across the data ecosystem. So civil society organisations have to be strengthened, so that, you know, they understand the data, they are able to use it, policymakers, academia, to do research, etc. So I think you really do need the resources to strengthen the whole ecosystem, so that I think everybody can meaningfully participate. I'm, again, really also focused on the issue of representation. I think it's critical, if we don't have the right people at the table in terms of data, data governance systems, we will have bad data that results. So where gender, gender equality specialists and advocates are not at the table, other issues will be addressed. So I really do think that it is important to make sure that the people, the right people are represented. And finally, I would say that, you know, currently, I think, you know, a lot of ah, and I think it is a bit spoke to that a little bit, there are, there's a lot of data that is being produced, there's a lot of work that is being supported by donors, and you know, UN agencies, other stakeholders, etc. But I'm afraid that this is not well coordinated, and it doesn't always respond to national priorities. So if even if we put a lot of resources on data, in terms of financial resources, and technical resources, if those do not respond to what countries have in their national plans, development plans and national statistical plans, then I think it becomes really efforts that are wasted, and that are dispersed as well. And that we cannot afford, I think there really needs to be a drive to respond to countries' priorities to make sure that we address what countries have in terms of what countries have expressed as their priorities, so that we can move forward as, in a more coordinated and efficacious way.

Deborah Hardoon: 24:12

So linking that to the point that you made about the need for investments, not just in the data collection, but in the infrastructure, do we need better financing mechanisms, and more coordination around the financing mechanisms? Because if the if the, the money and therefore the power all comes from donors or external agencies to support these statistical systems, maybe the there's some room for improvement there and how that's working?

Papa Seck: 24:41

Yeah, absolutely. I think, you know, again, better coordination and better financing is needed. So that, you know, again, responding to those priorities that have already been established by, you know, at national level, and that I think gets us far far enough. So what for instance, what we've been doing is really working with all stakeholders. In most of the countries wher we work to develop national gender statistics strategies. So these are, these are strategies that are defined by the entire ecosystem of stakeholders through these coordination mechanisms that I've mentioned. And we tried to call as all, you know, donors, partners, but also the government to invest in that, to implement that strategy. And that I think, again, gets us to be more coordinated, but also to make sure that all our resources are directed at that, you know, the same problems.

Deborah Hardoon: 25:37

Does this resonate in Uganda Elizabeth?

Elizabeth Birabwa Aliro: 25:39

It does resonate. But I wanted to emphasise that in addition to building the technical aspect, the infrastructure, the financing the resources, I think we also need to look at the soft and invisible aspects around evidence use around data ecosystem, especially in terms of culture shifts, we need to build trust relationships that are based on trust between all the actors that are involved in the data ecosystem, for them to be able to value and plan better and also resource how the data ecosystem is managed. I think this is something that is often overlooked, you find that the relationships have quite a big bearing on our success. If the generators, the brokers and users do not believe in a common agenda, it's very unlikely that we will succeed. So I want to emphasise that as much as we are focusing on investing in the hard technical issues, we should also look at the soft issues around relationships around organisational culture around the behaviours that influence and impact on evidence use, and data use.

Deborah Hardoon: 27:05

And what do you think we could do there?

Elizabeth Birabwa Aliro: 27:07

Ah for me, first and foremost, maybe something that is starting to happen in Uganda, where you see that both the state and non state actors seemingly coming together more to develop the national statistical system to develop strategies for gender disaggregated data. I think that is the kind of mode that we should see happening more unlike in the past, where the state actors have been taking this whole data agenda on their head, and excluding the non-state actors. But at least in the case of Uganda, we are seeing UN Women is really driving that kind of interaction, that kind of agenda of bringing together the different state and non-state actors, including those at the local levels, to come together and develop this whole gender disaggregated strategy making and programming. And we are also seeing that happening, as we are seeing the appreciation of the CSOs the civil society generated evidence where now guidelines are being developed, to see how they can be supported, to better generate data that will be mainstreamed within the overall national statistical system.

Papa Seck: 28:35

Just to add to that, I think, I recall, for instance, one conversation that we had during the Commission on the Status of Women here. And it was a side event that we organised that was really fascinating to me, where we had someone from civil society, head of a civil society organisation and a head of a national statistical office. And to summarise, basically, a 15 minute conversation went like this, your data is not good enough, and the other one saying your data is not useful to me. So and, you know, that was a 15 minute conversation. But I think, you know, again, if you take the case of Uganda, we really move really far from from from that oppositional conversation, where it's really trying to see, you know, what are the solutions that we can put in place so that we can really solve I think the problems that are coming from official data, where, you know, it's not it cannot be disaggregated. It doesn't address the proper points, to, also I think, on the other hand, helping civil society to develop essentially adhere to the standards that, so that the data can be used by official statisticians. So and that's really, I think, the kind of conversation again, that we that is needed and that we are really trying to push also around the world.

Deborah Hardoon: 29:53

Thank you both so much for your contributions in this conversation. I think there's been a really clear theme throughout this conversation that when we talk about data and data ecosystems to get that right, it's really not just a technocratic issue and something for national statistical offices to worry about or data nerds like like us to think about. But for them to work, we need really meaningful collaboration and inclusion with many different partners in the ecosystem, all the users, the people that this data is, you know, talking about, and so on. So I think that's been so appropriate given that this whole series of this podcast series is all about the leave no one behind agenda. And you know, the theme and conclusion from this conversation is that it's all about inclusion, collaboration and participation across the data process and the data lifecycle. So thank you so much for that. Before we go, I was just wondering if there was anything that either of you would like our listeners to take away to read or to think about as they digest this podcast?

Elizabeth Birabwa Aliro: 31:03

Yeah, what I would like to share is that, in Uganda, we were, we've just completed a diagnostic study, where we're trying to diagnose how evidence is used in planning, budgeting, and monitoring and evaluation. And this was mainly focusing on Uganda as it transited from sector-wide kind of programming to a programme based approach nature of programming. So we are looking at how evidence is being used. And we are using those three sectors. And that diagnostic report is out, it can be accessed via the EPRC website. But also other products of that study are also accessible on the EPRC website, which is WW dot EPRCug.org. Thank you.

Papa Seck: 32:07

Yeah, so for me, I mean, I will just add, so it won't be a commercial, although I will just ask, you know, if you can visit, if you are interested in any of the information that I've shared here, you can visit our website at data.unwomen.org, where you will find a lot of those resources. But again, for me, I think one thing is that, that I'm really finding to be quite important, as part of the work that I do is that data is important for all of us, right? So whether we can sit we think about it actively or not. So you know, and for all of those who are listening to these podcasts, get involved in data, understand what data is being used to, to make decisions about your lives, and make sure that this is the kind of right kind of data and this is particularly important today, given you know, fake news, etc. So really, I think, again, get involved in data and really recognise its importance in your life.

Elizabeth Birabwa Aliro: 33:08

Maybe the parting shot for me is we need to build an active citizenry, given the level of data privacy, or data security now in the digital age we are in, I think we need to[inaudible] the citizenry, so that the private sector, the state actors, don't take advantage of the so much data that they hold on behalf of the citizenry.

Deborah Hardoon: 33:40

Some really powerful closing thoughts there from our two speakers. Thank you so much. And thank you all for listening. We hope you found the discussion here interesting and useful, particularly in efforts to use data to leave no one behind. If you haven't already, I encourage you to tune in to our other two episodes in this series for more discussion on inequality, and on risk data in the context of the leave no one behind agenda. Good with data is a production of development initiatives and independent organisations that enables action through data driven evidence and insights to end poverty, reduce inequality and increase resilience. For more on what we do and the issues discussed in this episode, go to devinit.org that's d e v i n i t.org. This series is produced by Sarah Harries, Joshua Flynn, Anna Hope, Tim Molyneux and me, Deborah Hardoon.