 Good morning, everybody. Firstly, welcome to Cape Town. I can promise you there is a mountain that will probably appear in about two days' time. But my name is Haran Borat. I'm the director of the DPRU. And we're really, as the DPRU, the Development Policy Research Unit, like to welcome you to the sixth ISA World Bank NJD UNU-Wider Jobs and Development Conference. Sounds like the World Cup. All the sponsors are there. But the budget is much, much lower, I promise you. We struggled as a steering committee with many things, but including finding an identifying theme that would sound different and catchy. And we didn't really succeed. But what we do have is the sort of challenge of creating better jobs in developing countries is a catch-all. But I think what we do have is a community of, and I'll talk about this a little bit later, of niche operators that are really important in a developing country context. So I thought very, very quickly, let me give you perhaps for this community as I think most of us are aging in this sort of conference, know the history of this conference. But let me just step back. There was, I think it was 2013. Some of you may know her, Mary Holwood Dreamayer, who's from the World Bank, still at the World Bank in private sector development, I think, who came up with this idea of a network for jobs and development as part of a grant facility that she was putting a proposal together for. And what she did was to bring together institutions based in regions of the developing world, so Eastern Europe, East Asia, South Asia, Africa, Latin America, who were doing work on labor markets. And that was the niche she had identified, very, very important. So I don't mean it facetiously. We don't really want to research the impact of food prices on exchange rates, right? So we had development economists but actually we were working on labor markets. And she recognized that that particular area needed further support. And so she brought this community of us together and we did research for two to three years and the funding ran out, but the conference was the element within this jobs knowledge platform, as she called it, that continued. And so what you have is this network has stopped doing the research, if you like, but the conference lives on. And so in many ways, that's why you have the sixth conference that has now, over the years, included UNU wider and the World Bank and of course, ICA. And so in many ways, we'd like to think that the institution saw the value in continuing to fund this project. And so we've had conferences in Washington DC, in Bogota. We had the online conference and now you're here in Cape Town. We've saved the best for last, clearly. So just very, very quickly to come back to this idea, it struck me when I was thinking about, well, who are we as a community? I realize that it's almost a parallel process occurred in development economics, if you like, which is this growth of Haasol Survey, labor force survey data, census data over the last 15 years in developing countries. And so many of us as researchers working on developing countries may have taken a view that actually issues around wage inequality, around routine task intensity and its impact on the wage distribution, on employment equity, on discrimination are really important. And that's what I wanna do work on. So it became an area where developing country labor market research grew dramatically. And I think that's the niche we have here, right? We're not, I think it's inaccurate to call us maybe a bit controversial, development economists, right? We're not development economists. We economists working on developing country labor markets. And I think that's the niche. So if you look at the papers, and I've just run through some of them, I've gone through quite a few of them, you've got informal formal transitions amidst COVID-19 in Latin America, in Benin and in sub-Saharan Africa. You've got work on wage subsidies and youth employment in South Africa and in West Africa. You've got work on the impact of public works programs in fragile states in Africa. And if you wanted the interface between a developing country work and labor market work, that's it. Minimum wages and premature de-industrialization in Columbia. And so the list goes on. And we have a total of 50 odd papers brought together by I think you as the community. So this conference is very much yours. So with that, I'd like to open the conference but hand over to our fellow steering committees who are also gonna provide their introductory comments. I will then come back with the unfortunate sort of housekeeping rules and so on. But let me hand over to Gary. Should I stay here or go up? Please come up. Welcome, everybody. I have the distinct pleasure of speaking, I have the distinct pleasure of recording in progress of welcoming everybody institutionally on behalf of IZA. I want to begin by thanking the organizers. It's been marvelous job that you and your team have done Haroon and I'm not gonna at this time go through individual thank yous but just would not have been possible without the great support that you've provided. As you heard from Haroon, this is the sixth conference in this series involving the current sponsors. It actually goes back this conference does to 2006 when IZA and the World Bank teamed up on a jobs and development conference. The idea being to bring together people who were working on labor market issues, more and better jobs, how do labor markets function? What do we know about how they work well? What do we know about how they work badly? And to try to share lessons and experiences using the tools of labor economics. And so the current partnership is the latest incarnation. I have a feeling it's not gonna be the last one but there will be others who will be wanting to join as well. Now, the conference is set up to give a lot of time for people to give formal presentations but at the same time and now that we're back live which is a wonderful thing, we can actually do this. We can have these informal discussions with others about what we're doing, why we're doing it, what we're learning, what we haven't learned, what we hope to learn. And that's what makes it all wonderful. The most recent of which I just had at breakfast. And so I wanna welcome everybody for being here. Hope to meet all of you and to learn from you and thank you all for coming. Morning, everybody. My name's Ian Walker. I'm the manager in the jobs group at the World Bank. I just wanted to welcome everybody here. It's lovely to see everybody in person. In the now quite large number of years in which I've been manager in the jobs group, I think that helping to organize the jobs and developing development conferences, one of the most satisfying things that I get to work on. And it's satisfying because of the interface between the policy side of the problem which we tend to focus on in the bank and the academic studies which all of you do which provide us with inputs into thinking about better policies. I think personally I've come to think after a long time working on development that jobs really are beyond the sloganizing at the heart of the problem to the extent that we solve the problem of better jobs, we solve the problem of development and to the extent that we don't, I think that we fail. So this interface between labor economics and development economics, I think is at the heart of the profession now and I think it'll continue to be so. I think there's more and more awareness of that amongst people who've talked about growth for many years. They now talk explicitly about jobs and jobs transformations and there are lots of sessions at this conference which are in exactly that space. I'm very much looking forward to listening to the presentations and to discussing with all of you and also to chatting in the informal sessions. So welcome and thanks for coming. So good morning everyone. I'm Kunal Sen from UniWider. UniWider actually joined this on the set of Organisers in 2019. And I have to say that over this last few years, this conference has become the go-to conference for labor economics working on development and I might even say Haroon, perhaps for any development economist actually. It's just been amazing how good paper quality of the papers that we are getting, submissions. We have 64 papers in the panel sessions. We could easily selected 60 more papers. So that was really amazing. So the living panel had a tough time actually deciding on the papers and congratulations to all of you who are here today. I also want to say that the other thing I think which is very important for UniWider is that we're seeing over the years increasingly geographically diverse set of scholars presenting papers here. We have scholars from Africa. We have scholars from Asia, from Latin America and of course scholars from North America and from Europe. I do think that this conference for that reason is quite different from what we seem to see in development economics conferences in North America and Europe. And that's really important that we carry on this sort of very much inclusive approach to development economics because that's perhaps not as much evident sometimes in the Northern Hemisphere. So that's also very important for us. On the teams of the conference, UniWider has been working on quite a bit on informality and sort of change and gender and so on. So again, a lot of the issues the papers in this conference fits very well with the work we do in UniWider. I really encourage you to look at our websites and see our projects and see our publications. And so enjoy yourself. This is a fantastic venue for a conference and thanks so much to Haroon Sara Marit and of Sara Sara's here because she's really been behind it and others in from the DPR organizing team for putting on this fantastic fabulous show for us. I'm going to really enjoy ourselves the next two days. Thank you. Good morning. Good morning, everyone. My name is Pietro Van Doske. I'm the president of IBS which is the Eastern European partner of the network and on behalf of the network and the other partners who are actually not here this year, Institute for Emerging Market Studies at Hong Kong University of Science and Technology and Indian Council for Research on International Economic Relations. I would like to welcome all of you to the conference. As Kunal mentioned, it is truly a global conference. I've checked yesterday. We have speakers from more than 20 countries and all in habitat continents. So that's very unique in comparison to, as you mentioned, many conferences. Each year we get about, we'll get more than 300 papers submitted. So really appreciate your interest in the event, your time and effort. And we were supposed to host the conference in Warsaw in 2020. It didn't happen for obvious reasons. Then we had another online event in 2021 which was very cool that the exchange was able to continue during the pandemic. But I think that for all of us researchers, policymakers, that's actually one of the reasons why we do these jobs to have this kind of exchange in person. So be nice to each other, talk to each other, take advantage of the breaks of the dinners because quite often what we remember from this kind of conferences are the papers, but also the contact, the new friendships that we make, what we learn in this kind of informal setting. So thanks so much, Harun, for hosting all of us here and I'll give it back to you. Thanks, Piotre. So the wedding speeches are over, but we sort of had to do it. Welcome you as, so you can see the team that's been doing all the work. And just to add that we spend a lot of time including the DPRU researchers, many of whom you'll meet over the next few days. Looking at the papers and the submissions. So it was incredibly difficult to make the choices that we did, as Kunal said. So well done actually for being one of the presenters at the conference. I have a few housekeeping rules, which I'm really bad at usually, but I run through them. I think these are known to all of you, keep the mobiles and the cell phones off. You're not undergraduates. So I know you won't be talking on your phone while we lecturing and people are talking. Health and safety issues, please just go to the front desk and they will guide you for any health issues. There are restrooms on every floor, so that's fine. So in case like me, you are confused, we're down here. We are upstairs for the parallel sessions. And then the food is on the roof terrace, level four, right? And you'll see the DPRU staff around at the conference. So please ask them if you're confused about anything and Mira, Wadia, Kezia, Janine are all around as the administrators. So maybe put up your hands so people can see who you are in case you're confused. And then we've given you all Apple Tag, so we can make sure you don't go to the waterfront during the sessions. Okay, what we're going to do is hopefully have our Zoom session ready. And the idea was to avoid getting a minister because the ministers always, those of you in developing countries know, they'll always at the last minute say, oops, sorry, I can't make it. So we actually got one better. We got the economic advisor to the president, Trudy Makaya to give us opening remarks. I want to check with people if Trudy is ready for the Zoom so I can introduce her while she's actually online, taking guidance. And will she come up on the screen? Here we go. Good morning, Trudy. Can you hear us? Good morning, yes, I can hear you. Fantastic, Trudy. I'm just going to introduce you and then you have a wonderful audience, all facing you on a big screen. Hopefully you can see some of us. But just for our audience, Trudy is the full-time economic advisor to his Excellency President, Sororoma Posa. She provides technical support to the president on economic policy, including regular input on key issues and initiatives, interfacing with relevant advisory councils and engaging with economic policy stakeholders. I know Trudy very well. She's incredibly focused and competent and as part of our economic advisory council to the president has been a very, very important instrument in the bridge to working with the president and cabinet. Prior to that, Trudy was CEO of Makaya Advisory, which was a consulting firm with a focus on helping business navigate economic policy, including competition policy and which he founded in 2015. She's also served as an advisor and investor to young companies and has held non-executive directorships for Vimalana Advisory Fund and MTN South Africa, which is a telecoms company. She holds an MBA and an MSc in development economics from Oxford University. So Trudy, over to you, but welcome to this conference and very, very grateful that you found the time to open it for us. Thank you. Good morning, program director. Thank you for that very comprehensive and kind introduction. And good morning to esteemed delegates. I'm really sorry that I couldn't be with you in person, but other commitments have kept me in counting, but I'm really glad to be able to open this conference. This November last month, when the G20 held its summit in Bali, the global community wondered if the leaders of this economically significant block would be able to issue a joint declaration based on consensus by all 20 members. The relevance of the G20 as the premier forum for global economic cooperation was in doubt. And when against most predictions, a consensus declaration was delivered, analysts poured over the language on the armed conflict in Ukraine. But this morning, I would like to take you to a seemingly uncontroversial, less scrutinized sentence in this declaration. And it goes like this, I quote, in line with the Indonesian G20 presidency theme, recover together, recover stronger. We will take coordinated actions to advance an agenda for a strong, inclusive and resilient global recovery and sustainable development that delivers jobs and growth, jobs and growth. Even though jobless growth is not entirely an accurate characterization of the South African economy in the 2000s, growth has not been synonymous with job creation in our economy. Hence our insistence at the G20 that the leader statement should not only talk to growth but also to job creation. Indeed, across the world, the post pandemic recovery threatens to yield very little in the way of job creation. Recovering the jobs that have been lost is proven to be challenging, let alone creating new jobs about the 2019 baseline. The G20 leader's letteration went on to say, the rise of automation and digital technologies are reshaping the world of work, presenting both opportunities and challenges. Adding to the situation, the COVID-19 pandemic has exacerbated pre-existing inequalities in many countries and continues to disproportionately affect women, youth, older workers, persons with disabilities and migrant workers. We underline that it remains our utmost priority to mitigate the adverse impact of the current trends on the labor market, reduce inequalities while responding effectively to the opportunities that automation and digital technologies present and promote gender equality. We remain committed to the promotion of decent work and the elimination of child enforced labor. There's also other language on the importance of community-based vocational education and training on the empowerment of entrepreneurship, particularly through MSMEs and also the importance to extend labor protection for all workers, including those in the informal sector. So we see a very solid encapsulation of the aspirations of the G20 and how leaders, even with many other priorities on the table, are still focused on the importance of job creation and of dealing with those pre-existing inequalities and labor markets, while also thinking about how to deal with the threats of digitalization and other forces in the economy. I think the statement is also a reminder that though the G20 includes emerging economies, even advanced economies have to think about developmental issues in labor markets, be it the lack of inclusion or displacement through globalization or thinking about where the jobs of the future are going to come from. Now, if we look at the challenge in South Africa, there are various explanations of the high rates of unemployment that policy makers have to grapple with. They've been summarized and diagnosed in many reports, but here I will quote the high level panel report that was published by parliament in 2016, which kind of frames it nicely in terms of looking at the various explanations for why we have the unemployment outcomes that we do have. And these, according to the report, include the sectoral structure of the economy which is not sufficiently labor intensive. Further, the agricultural sector is unusually small for a developing economy and deindustrialization has led to a decline of employment in manufacturing. There's also the concentration of productive activity in a few firms within sectors that limits the potential growth of small businesses and new entrants into those sectors. The spatial distribution of economic activity is skewed with a large number of people living far from centers of work, in a sense reproducing the spatial legacy that we inherited from a Part 8. The report also pointed to excessive and bureaucratic regulation of business activity and labor markets, which stifles job creation, especially for small businesses. It has been argued that if South Africa were to match the self-employment rate of its developing economy peers, the unemployment rate would be significantly lower. And we saw earlier this year in President Ramaphosa's State of the Nation address the announcement of a serious initiative to tackle red tape, which stifles business in general, but is particularly biting when it comes to small businesses and the informal sector. The panel report, in terms of expanding the high unemployment rates, also looked at high transaction costs and inefficient network industries which reduce competitiveness and limit export growth and finally insufficient and poorly met skills, which means that South Africa has a low-skill space while the structure of the economy is biased towards skill-intensive activities. So there are fewer jobs for the skills profile that we actually have. The skills issue in particular is not a question of resources because in both the public sector and in private institution, it has been shown that South Africa puts in a lot of money comparable to its peers, sometimes even ahead of its peers and yet showing very little in terms of outcomes for that investment. And so we constantly have to think about why resource allocation is not yielding those results and some of it might be related to poor coordination of skills development, the lack of capacity within some of our technical, vocational education and training colleges and also asset to education and training authorities which are the entities that really should be providing the core of the vocational skills and job-related skills that we need. However, they have been promising partnerships between companies and tivets, including with multinationals that come from countries with stronger systems of apprenticeship and training. But initiatives such as these need to drive fundamental changes within the skills development ecosystem. We understand that economic research and policy should provide a clear and comprehensive answer to the plight of the 11 million and counting adults who are unemployed. And we found that too many of our proposals and interventions in the past have not operated at the right scale. So we have many programs that deal with 200 beneficiaries here, 1,000 people trained there whereas we actually need something that works at scale. So it is encouraging that we have started to tackle that issue and that during this conference you will hear from some of my colleagues in government who have worked on programs that drive for scale such as the presidential employment interventions and also those programs that seek to provide a demonstration effect for systemic change such as the jobs fund. If I now look at the current policy stance in government is one of pursuing a multi-pronged strategy towards dealing with unemployment. This recognizes that the unemployed are diverse and that various segments of this demographic will require distinct solutions. So there's no magic bullet for these millions of people who are unemployed or who stand outside the labor market. For some of the unemployed work will come from the private sector. And so this involves deepening the work that we've done in terms of economic reform. If we think about the diagnosis for our unemployment crisis we talked about the inefficiency in network industries and high transaction costs. And so our economic reform programs are geared towards unlocking potential growth and also removing barriers to economic activity. The most prominent expression of this reform program is Operation Vulinjela, which is a joint program by presidency and national treasury that spends across government to look at various areas where reform is required, including in rail, energy, telecoms, the skills regime and other priorities that are added as the program matures. Now, as I mentioned, there's no magic bullet. So for some of the unemployed, especially in the short term, work will come from publicly funded employment. This has included our reimagining of public employment programs as was done in the economic reconstruction and recovery plans, presidential employment stimulus. This was one of the most significant and rapidly scaled up public employment programs in the nation's history. And it's quite remarkable that this was achieved during the pandemic, during lockdowns, when coordination would not necessarily have been seamless, but government was able to come together to deliver such an important public employment program. We continue to think about how to scale up meaningful public employment that creates long-term value for communities. We know that in the past, there have been many criticisms of public works programs across the world in terms of whether they actually prepare their beneficiaries for long-term employment and an exit into more sustainable forms of livelihood. And so we're thinking carefully about how we show that our public employment programs achieve this. We would have seen in the presidential employment stimulus also that a lot of the jobs opportunities that were created were in fields such as teaching assistants in schools, environmental programs, digitization programs. So where one can see real impact in terms of the work that's been created and also hard and soft skills development for the beneficiaries and participants in those programs. The state also supports private sector employment. So publicly funded employment, but supporting the private sector through programs funded by the fiscals like the employment tax incentive to encourage youth employment. To yet another segment of the unemployed, elevating and supporting individual and community efforts to boost informal and small enterprise is the answer. We're working to improve linkages between governments, anti-poverty strategies and the creation of sustainable livelihoods at community level. And finally, for those working age adults unable to find productive employment, social protection will continue to be offered on a fiscally prudent scale, building on the seeds of the special COVID-19 social relief of distress grant. Further policy development will explore options for using social protection as a vehicle to strengthen labor force participation. Even if accessing those social protections, labor force participation is not a condition of access to programs, but they would be designed in such a way that there is complementarity between social protection and job search. So we have crafted this multi prompt approach to try and achieve a structural break in labor market outcomes. And necessarily this will involve all of these four pillars. Private sector employment in a growing economy, publicly funded employment, social protection and enabling sustainable livelihoods. There's no point in elevating one above the other or reducing policy to just one approach because these have to work together and address the distinct needs of the various segments of the unemployed. Now, when we look at future dynamics, we know that the green and the digital transitions in how societies function have been held up as threats to job creation or as massive opportunities for job creation depending on who you talk to. This year in South Africa, we saw the successful conclusion of the auction of radio frequency spectrum, a key enabler of the digital economy. The Presidential Fourth IR Commission, Fourth Industrial Revolution Commission proposed a range of programs to develop all facets of the digital economy, including through skills development in partnership with the private sector, infrastructure development for the digital economy, regulatory reform and research and development. On the climate front, we have also been accelerating our climate action in the context of a just transition and sustainable development. Since COP26, a just transition framework has been finalized with the support of the Presidential Climate Change, Presidential Climate Commission. At COP27, just weeks ago, we presented the investment plan to support our just energy transition highlighting three key areas, renewable energy, the hydrogen economy and the transitioning of our automotive sector to new energy vehicles. So we very much on the opportunity and optimistic side of seeing these transitions, but it's important that these value chains in the digital economy, but also in the emerging sectors of the green economy are supported and invested in in a way that ensures that the jobs that we lose from other parts of the economy that are no longer relevant in this new world that we can compensate, more than compensate for those jobs and for those skills that we need and that communities do not bear a disproportionate impact of those transitions. Speaking in his role as the chair of the ILO Global Commission on the future of work, which launched its report in 2019, President Zoramacosa warned that a period of global cooperation was giving way to unilateralism and narrow nationalism. He went further to trace an element of this turn inward to the outcomes of labor markets, whether they deliver inclusive and equitable outcomes he said, or whether they reinforce inequality and amplify uncertainty. And to the extent that they do the letter, we will see fragmentation in the global economy. The recommendations of that ILO report to remain as relevant as ever, increasing investment in people's capabilities, particularly those needed to thrive in a carbon neutral digital age, increased investment in the institutions of work, such as employment contracts and regulations, minimum wage dispensations and labor inspection systems, social protection and many others, which are the building blocks of just societies and increasing investment and decent and sustainable work. In conclusion, I would like to leave you with the message that South Africa is on a path towards a sustained recovery. Recent economic indicators have shown a decline in the unemployment rate by a percentage point to 32.9%, though it remains too high and better than expected quarterly GDP growth quarterly at 1.6%. Yet we remain conscious of the fact that deliberate and focused policy interventions towards job creation and sustainable livelihoods have to occupy a central role in economic policymaking. There remain many difficult and uncomfortable questions for research and scholarship to share a light one. For example, does job creation in the South African economy possibly mean the expansion of low productivity activities instead of the high value added scale intensive activities that we hanker for? Is it inevitable that in the short term, economic growth will reinforce the minerals energy complex albeit in a greener and more sustainable direction, but in so doing remaining capital intensive and jobs deficient or more optimistically, could we unleash job creation and sustainable livelihoods from the informal sector? But bearing in mind that perhaps the average citizen might not conceptualize these as real jobs. I will leave you with those questions and I thank you all the best with your conference. Thank you very, very much, Trudy. You're welcome to stay, of course. Before we hand over to Gary to lead us in the opening address, Jadisi Oriana has arrived. I was looking for her. If you, if I'm hoping that you use the lens of the country that you've worked on or countries that you've worked on or live in, right? There were themes that resonate. Even though this was South Africa focused, you heard premature deindustrialization. You heard regulatory barriers to product and factor markets, skills bias technical chains, skills mismatch and the quality of schooling, public works programs and other forms of active labor market interventions, social protection, sort of with conditionalities, right? But maybe, right? That's a huge issue, but again, very common theme. And then finally, looking at Ian who's here as well, green jobs, but let's not forget the just transition. And I think not only are all of those themes actually covered in the conference, but they probably resonate with many of you in the countries that you work in. So I'm actually really glad that Trudy, by example, led with the South Africa specific intervention, but it carries so many sort of common thematic areas of investigation for many of you. So I'm gonna pause here and hand over immediately to Gary to introduce our opening speaker, Gary. We're gonna go for a full hour now. Okay, welcome again, everybody. It's my task and my pleasure to introduce the first of our keynote speakers, Oriana Bandiera. I'm not going to speak for long. I just wanna say that she has a remarkable, extraordinary record and is one of the most distinguished people in our field. She holds the Sir Anthony Atkinson Chair in Economics at the London School of Economics. For those of us who work in the field of inequality, this is one of the absolute greats of all time working in that field and it's a real honor for her to have that position. Other positions she has, she's co-editor of Econometrica. She's the vice president of the European Economic Association. She has a CV that we talk about publishing in top fives. And one paper maybe, two papers maybe, but she has all of her papers, nearly all of them in top five journals and a remarkably long list. Very simply, she is one of the brightest stars in our field and we are deeply honored to have her here today. She's going to speak to us about a new initiative that she is the director of. She is the founding director of something with a very long acronym, but what the acronym stands for is gender growth and labor markets in low-income countries. And the session that was held yesterday represented some of the outputs of that new program. What she's going to speak to us about today is one of G2LM-LICs projects, which is the title of her talk. It is economic development and the organization of labor, evidence from the jobs of the world project. Oriana, please come forward. Thank you, Gary, for this lovely introduction. Thank you. I was only seeing myself on the screen, I was a bit worried. So thank you so much for the organizer. This is a beautiful place and you win by far the best badge linear that I've ever seen. And this project actually follows quite neatly from the introductory remark that the economic advisor gave. I don't need to preach to this crowd that labor is the most important thing. I need to do this for other audiences, but we do know that if we think of poverty, the only thing that poor people have by definition is labor. But not only at the macro level, labor is the main factor of production in all economies. For all that we talk about capital and skills and new machinery and a digital economy, in the end, it all comes down to people. So understanding whether labor is allocated efficiently is key to understand both poverty and differences in national income. Now, what I propose in this paper, in this body of work, is that the organization of labor that is the type of jobs and their allocation, actually the disparity in the allocation of labor along lines of gender or race or religion anything that's orthogonal to labor skills, create the link between individual poverty and national income. So we have to understand how labor and jobs are allocated to understand both poverty and national income. The micro and the macro are connected and it is inequality that connects them. So I'm gonna introduce a project which ISDA and FCDO, the Foreign Office and Development Office of the UK government have kindly let me do, it seemed a bit madness when I first proposed it, but I think most of what I do does seem a bit mad when I first proposed it. But my idea was that, we do a million randomized control trials, individual level interventions of giving skills, giving capital, giving skills and capital, giving early childhood development, giving this, giving that. But these are always very location specific. And they're always embodied in a labor market, but we don't know how representative those labor markets are of the country as a whole or the world as a whole. So when I first started working with BRAC in Bangladesh, we started writing the questionnaire for the ultra poor evaluation. We had to make a list of all the jobs that could possibly be done in villages in rural Bangladesh. And we made a list of 99 jobs, took us forever, there were 99. Once we did the survey, the women in those villages were doing three of these 99 jobs. Everywhere in all the villages, we have 1600 villages in that sample. Everywhere there were three jobs at most. Was that normal? Is there something like the way we asked the question? So I started thinking more and more about how we can provide micro evidence at a macro level. And that's where I came with the idea of putting together all the public resources that we have, all the survey that we have to create a micro database at the cross country level. And this is what came out of it. We have 115 countries across the development spectrum. And we build comparable wealth quintile measures using the assets. And this basically gives us a way of merging the databases. And we can look at macro statistics on labor markets using micro data. The data can be downloaded from the website and also the codes so you can customize the codes as you like. This is the coverage. And you see the light blue is the DHS. And the dark blue is the IPOMS, the census data. You see that combining the two allows you to cover almost all of sub-Saharan Africa. There's only a bit of North Africa which is not covered by any of the two sources. And in previous attempts at using micro data to examine macro issues, normally there would be one or two African countries because the IPOMS doesn't cover them. But by combining the two, we can cover everything. So let's start with the definition of work. Again, preaching to the choir here, so I don't need to go in great details about this. But I just want to make one point. So ideally, work we would want to measure any activity that creates value and that can be done by others. That is everything but sleep, which unfortunately I've been trying to contract out, but I've never managed to. In reality, both the DHS and the IPOMS follow the system of national accounts. That is, they look at production rather than productivity. And they count it regardless of whether it's for sale or on consumption. This creates a lot of inconsistency. For instance, if you cultivate tomatoes for consuming them, that counts as work. If you use the tomatoes to make tomato sauce at home, the labor that goes into making the tomato sauce does not count as work. So what we call work loosely is effectively measured work. And there are lots of activities, and in particular, services that are provided within the household that are not part of this definition. And the reason I'm pointing this out is that it typically leads to an official underestimate of women's contribution to the labor force. And I will conclude coming back to this point of women participation and the contribution of women to the economy. But looking at a measure of work that excludes the majority of activities that women do, because that doesn't make things particularly easy. That's not me. So I'm going to talk about how jobs vary along the process of development. And I'm going to talk about three transitions that happen. The first is the marketization of work, where we go from home production to market transactions. Now, none of these transitions happen suddenly. They happen gradually so that within a given country, you have a bit of everything. So we will also see how they spread across time and within a country and how they spread by gender and wealth. So if you start from the most basic economic systems, people consume what they produce. Everybody has a subsistence level. So work is unpaid by definition because there's no market where to exchange work. You go from that to a system of market where you sell what you produce and you buy other things in returns. This is from unpaid to paid work. The second big, so this is when markets arrive, the second big change is the emergence of firms as the main organizing unit of work. And this is the movement from self-employment to wage work. It's very important to distinguish self-employment activities which are done because there's nothing else to do and self-employment activities which are done because somebody has a brilliant idea and wants to start a business. A lot of programs which are meant to promote self-employment tend to improve the lot of every existing little activity that there is on the market. But many of these activities are there because the owners have no better job to do. So we have a very large number of interventions which try to grow businesses which the owners do not want to grow. And so all this, see you nodding vigorously. Not surprisingly, these programs do not work very well in their intended state. I always wonder why before trying a program we don't ask people, what would you like? It would save us quite a bit of money, but anyway. Then I will move from that to the movement within wage work and how wage work evolves. So how the growth of firms leads to specialization and the creation of new jobs within the firm. And that will be most of what I will talk about today. So marketization of work. This graph which is not labeled, I'm sorry, shows the share of people in employment along the GDP path. So this is log GDP. It goes, say from Ethiopia in 2000. I think it's the lowest point and Switzerland in 2010 or so is the highest point. And what you see the tick line is the share of people in employment. The dashed line at the bottom is the share of people in unpaid employment. And the dotted line at the top is the share of people in paid employment. So you see that there is kind of a mild U-shaped curve because the paid employment grows less quickly than the unpaid employment declines. But this is very gender specific. Now this graph splits it by women and men, women in orange, men in purple. The color coding took a while to choose. We wanted to come up with the least gender specific color coding possible. So it came out in this kind of 70s vibe of orange and purple. And you see that the U-shape is mostly driven by women as we will know. The labor force participation of women relative to development has a U-shape pattern whereas for men it seems to be decreasing all alone. Now, when we look at the marketization of work by wealth and gender, we see that we see many interesting patterns. On the left hand side graph, I have the share in paid work overall. And the different shades of blue are the different quintiles of wealth. So the darker blue is the highest quintile, the wealthier people in the country. And the lighter blue is the poorest people in the country. And you see how it's monotonic. So it goes down in gradation of blue, meaning that the very richest are the first ones to get into paid work followed by the second quintile, the third quintile up to the fifth quintile. And the fifth quintile gets into paid work very late in terms of GDP space. Now wealth matters, but gender matters a lot more. The graph on the right has the same variables split by men and women. Men again are always in purple and women in orange. And you see that the first ones to get paid jobs are wealthy men followed by less wealthy men, less wealthy men and so on. And then eventually women. And you see that the U-shape is driven mostly by the poorest women. I should say that, actually I forgot to say at the beginning, this is joint work with a lot of people and all the beautiful graphs that you see are due to my co-authors, not to me. And actually one of my great co-authors is here in the audience, Hamid. So we can, and actually one of the great engines behind this big data set that we put together. So always grateful to my co-authors for all this work. So that was the first transition from paid, so from home production to markets. The second big transition is what happens on the emergence of firms. When people come together in an organization to put together their skills so that the total is more than the sums of the part. Now, the red line is the share of people who are self-employed. The blue line is the share of people who have a way job. That again tells you that there's no rich economy which is mostly made by small firms. Development comes by putting people together in firms. Now, this is actually, as we know, there is a process of structural transformation that goes on with development. So you might think that all that we are capturing here is just people moving from the village to the city and moving from small businesses in agriculture to firms in the city. But actually this is a process that works independently of structural transformation. You see that both in urban areas and in rural areas you have that decrease in self-employment and increasing wage job. So the transformation in the organization of labor happens in every sector. So it's kind of orthogonal to the sectoral change. We split it again by gender and wealth. So who gets wage jobs first? And by now you probably have guessed it is rich men that start with the wage jobs. So at every level of development, again these graphs have the same structure as the ones earlier. You have on the left-hand side graph the share of population in wage work by quintiles of wealth. Dark blue is the wealthiest, light blue is the poorest. And you see there is degradation that the rich go first and then go down. And if you split it by men and women it's not quite as separate as the paid versus unpaid work was but it's rather strikingly rich men get wage jobs first followed by poorer men and then eventually women. Now this I think is very important when we talk about the status of different jobs. So imagine in any society where the rich people are the ones who have wage jobs, wage jobs will have greater status. So when we think, I remember the discussion that we had I had like I arrived in South Africa and we start talking about unemployment and why people don't take wage jobs. And then I had that discussion a lot of times with people in the last two days there is a very high rate of unemployment in South Africa as we just heard as well from the advisor. And yet people do not take easily available self-employment activities like small trading that they could do easily. That's only done by migrants in the country. And maybe this is one reason why because if wage jobs have the higher status and being self-employed is considered not such a high status occupation people might prefer to wait until they find a good wage job rather than engage in activities that they find less status worthy. So it's important to keep in mind when we try to understand human behavior that we're not just motivated by money and having any form of income. The status that's associated with that occupation is also quite important. So what about informality? So here I'm just talking about paid versus unpaid wage versus self-employment. Now, when you want to make comparisons across the spectrum of development you kind of like discipline by the fact that you need to use variables that are comparable throughout. Informality, so paid versus unpaid and wage versus self-employment are objective definitions that are defined in the same way everywhere. You either get paid or you don't. You either work on your own account or you're hired by somebody. There's no ambiguity there. Formality however is defined differently in different settings and there is no uniformly accepted definition of formality. Sometimes it has to do with the workers that you employ whether they have a formal contract or not. Sometimes it has to do with the fact that you registered with the authorities, with the state or not. And in many cases it reflects state capacity more than the nature of work itself. So that's why there's no formality versus informality here. The self-employment largely reflects that type of informality that we normally think about when we talk about informality in low income countries. And another big missing thing is unemployment from the statistics. Why haven't I showed you statistics on unemployment? Because to define unemployment, we need employment to start with, right? So how do we define unemployment? Are you willing but unable to find a job at the going wage rate? So you need there to be a wage rate. That's an equilibrium wage rate that you're willing to work at. But when labor markets are very thin and wage jobs are very scarce, this is not defined. So we cannot really compare across the development spectrum. So this got me thinking actually also listening to the opening remarks of how we compare, you know, South Africa is a 30% unemployment rate we heard. How does that compare to an employment rate of Zambia, for instance, neighboring country? So Zambia has a lower unemployment rate, but how is it measured, right? So in a country where most people are self-employed, what does a low employment rate mean? Mostly if you have lots of self-employed, they would be underemployed because they don't work all the hours that probably they would wish to work. So these are very difficult comparisons to make across the development spectrum because it makes sense to define unemployment only if you have wage employment to start with. Nevertheless, I think it's a very important topic to study. And it might explain why wage work grows very quickly against GDP. So we don't have many countries where you have half of the people in wage work and half of the people in self-employment. You either have most people in self-employment or very rapidly you go to most people in wage work. So this is food for thought. I don't have an answer about it. What I want to focus on in the time that I have left, which time do I have? 20? Okay, fantastic. I want to focus on the third transition, which is the transition from small firms to large firms and what that means for jobs. So with development, firms grow. We know that we go from this small self-employee to a few entrepreneurs leading large firms hiring many workers. And what we know that when firms grow, layers form because one entrepreneurs cannot possibly manage many workers. They start developing layers. You have layers of management, layer upon layer. Now, is that all? This project came out actually of when we were putting together the job of the world project and we were looking at the codes for work, for labor and for the different jobs that people do in different countries and they looked completely different. Not just in terms of values, but in terms of numbers of jobs that were available. So we started thinking of how to make sense of this. And if you think from the point of view of a firm, jobs, you can think of them as a sequence of tasks, right? So a job is a combination of things that you have to do. Like the job of a professor is to do research, to teach, to get grants, to manage those grants, sometimes to book their travel, okay? And there is a link between productivity and task specialization, okay? One production function and tasks. You can have the person doing all the tasks, so jackable trades, or each person does one task and that leads to higher productivity. Now, of course, Adam Smith said this, so it's not particularly new. But where Adam Smith was focusing on is the learning by doing that you get if you do always the same task. So if I only book travel, I would become really efficient at it, I doubt it. But that's the idea, right? You always do the same thing. Now the same nail, you become quicker at it. But there is another important mechanism that leads to higher productivity when you have more differentiated jobs. And that is that you can match people by skills, preferences and jobs better, okay? Now, this is very different, I think from what people have in mind when they think about the vertical differentiation of jobs. We talk a lot about the allocation of talent, as if talent were a vertical measure of people's goodness in a way. And we talk about putting more talented people at the top because when they are at the top, they can influence many of the less talented people. But if you think about talent, one dimension of talent, what is this exactly? Can you pinpoint in a person who makes that person more or less talented? All of us are good at something and not good at something else, okay? So I'm very good at playing Tetris but hardly anything else. So people were good at maths and some people are wonderful writers, people were wonderful painters, people were wonderful cooks. A lot of the differentiation across people is not really vertical, it's horizontal. And many papers that have been written on the differentiation of labor are on the vertical dimension. Maybe because economics is a very hierarchical discipline, we only think of putting people in a line, you know, who goes at the top, who stays at the bottom. And we kind of fail to consider this horizontal dimension, both in terms of skills and most importantly in terms of preferences. If you think about it, you get jobs well done when the people who are doing them love what they do, right? So you get, you know, kindergarten, which are run fantastically well by people who are incredibly good with children because they like doing that. The same person with the same skills but different preferences would be a disaster in that role. So this is a long story to say that the more you can split jobs, the more you can split the tasks in different jobs, the better chance you have to match people with the job that they are better at and that they enjoy doing more, okay? So we expect a two-way relationship, the number of jobs to productivity because of the quality of the match and because of learning by doing and then from productivity to the number of jobs because you need, because there might be returns to scale, so you have to pay a fixed cost to split a job in different bits as well as technology and the product complexity that you're producing might determine the extent to which you can split jobs. Now, how do we measure job variety? This took us surprisingly long because what we need is a measure that does not depend on the size of the population or the subpopulation that we're looking at because we wanna look at job variety along different dimensions, say gender, say wealth class and so on, okay? So we want a measure that increases in the number of jobs, sorry, a measure of jobs variety that increases in job variety makes little sense. We need a measure that doesn't depend on the size of the population that increases in the number of jobs and that captures horizontal specialization as well as the vertical differentiation. So IPOMS records all the occupations that are based on the raw occupation variables used by each census bureau and the ISCO 88 classification, I guess everybody here is familiar with it, has nine major groups and 116 possible minor groups at the three-digit level. So our measure is the minimum number of jobs that accounts for 90% of the workforce in a given population of interest. Let me give you an example. This is Mozambique in 2007. These are all the jobs that account for 90% of the workforce. So there are about 13 jobs. This is what accounts for most of the workforce in Mozambique. As you go up the development spectrum, this is India, see the number becomes bigger. Within each category, it becomes bigger too. So you don't just add on the vertical axis, you also add within ISCO group and this is France. Okay. So job variety grows with development across country. So this is number of jobs accounting for 90% of the workforce and log GDP per capita on the horizontal axis. And interestingly, this took a fair amount of work. This is the oldest census data that we have from 1850 to 1910 for the UK, Scandinavia, Canada, and the US. And you see that there is a very similar evolution over time against as these countries grow. Now, a few colorful graphs to wake you up. This is by ISCO one digit, okay? The green is agriculture. And you have the agriculture strings and it's substituted by elementary occupations, machine operators and crafts and then from the top by professional clerks services and so on. Now this is gonna get quite involved. I'll build it little by little. This is countries, okay? All the countries in the sample by job frequency. So if you start from the poorest country in our sample is Ethiopia, you see that there are four jobs that account for 90% of the workforce. And the length of the bar is the frequency of that job. So there is one job that accounts for most of it. Now, we add the ISCO one digit color codes and you see here that that job in Ethiopia is agriculture. Now, this graph contains a fair amount of information because the darkness of each color represents the skill content of that job, okay? Which we measure by the average level, average years of schooling of the people engaged in that job, okay? So you see that there's two things that happen with development. You move away obviously from agriculture and you move from as we saw from bottom and top but also the lines within each color become more fragmented. I don't know if you can see well from the back but there are white lines in each of those bars. Each white line represents a new job. So the lines as you move from left to right in this graph there are more and more white lines. And the colors become darker quicker which means that there are more dark colors more skill intensive jobs. So we see here this transformation which is a larger number of jobs for every ISCO one. So there is not just a vertical differentiation of jobs but within each ISCO code and horizontal differentiation too, right? Now, so what? This is beautiful descriptive statistics still descriptive statistics. It's still cross country variation, okay? So what we're gonna do in eventually when we'll have a paper is to zoom in in a country with sufficient local diversity and exploit the within industry region variation, okay? And we'll use demand shocks to identify tipping points between different composition of jobs if there are any. The country is Brazil and the data that I'm gonna show you now are between the censuses in 2000 and 2010. There are 59 industries at the 2DG level and 558 micro regions. There is huge variation in development levels in Brazil. That's why we chose Brazil for this exercise. Now, these graphs basically show you how Brazil in different years compares to the countries that we have in the cross country sample. I'm not gonna go into great details just to illustrate that there is a very broad spectrum of development in Brazil. And that's a very good idea to use, that's why it's a good idea to use Brazil. Now, we did that graph. You saw it at the cross country level, job variety and development. This is how it looks across micro regions in Brazil in 2010, okay? It's fairly similar. That's the graphs that you saw before by frequency. This is done by the style of GDP per capita of the different micro regions in Brazil. You see the same pattern. Jobs become more diverse starting from the top all the way to the bottom. These are by the same thing, ESCO-1 frequency and skill intensity. And as you can see again, agriculture disappears and within ESCO-1 there is more variety of jobs, more on the skill side. And now this is what we can do with Brazil which we cannot do at the cross country level which is to look at specific industries and use that variation. So use variation at the industry micro region level. So you can throw in a lot of fixed effects and use only small variation within each of the cells. And you see that industries that have bigger have more job variety. Now, because we have two years a clean in the census we'll have five eventually, you can also look at the relationship with growth. So industries that grow more have more job variety, have job variety that grows more. And finally, this is actually, this requires a bit of explanation. We can look at the wage experience profile. There is a recent findings by my colleague Ben Moll and co-authors that show how in low income countries the wage experience profile is much flatter. So there is no much chance for people to grow and have a career than in higher income countries. What we do here is the same exercise. So these are simply estimates of the returns to experience. Of to each person, this is only for men on early wages. And the three lines are split by terciles of the job variety measures, okay? So the blue line is in the region and industries which are at the bottom tercile of the job variety measure. The green line is at the middle and the red line is at the top. And what you see is that as a first approximation of a better match in regions and industries where there is more job variety, wages increase more with experience. This is all unconditional. So take it with a pinch of salt. We will do it properly in due course. So to conclude, I want to, this is the summary you've seen. I want to talk about two issues. The first is redistribution, who benefits from this? And the second is how do we interpret it? Where is the direction of causality? Is it a loop? And in that case, can there be multiple equilibria? So first thing first, job variety is correlated with average income. Most importantly is negatively correlated with inequality. So it seems like to a certain extent, the increasing job variety and the better matches that ensue lower inequality. However, once we look at different types of people, so that's inequality earned income, if we look by, sorry, for the typo, if we look by race in Brazil, you see that there is more job variety for whites than for non-whites, and a lot more for men than for women. The white bits there are people who are out of the labor force, okay? So women to a much larger extent are out of the labor force. This is true also across countries, although across countries the relationship with GDP is not quite as linear as we know. So that brings us to check whether the occupational segregation, that is the difference in the jobs that men and women do increases with job variety, because that can be a consequence if you have many more jobs, then you can split them by gender more easily. And so it looks like that's the case. So as you have more job variety, there is more occupational segregation. So why do we care about this? Well, first is a question of justice, right? Gender neutrality is both intrinsic and instrumental benefits. The intrinsic value again is due to the fact that social status, educational and economic opportunities and political power are all closed link to paid market work. So as long as unpaid home production is not afforded the same status, access to jobs is a question of distributive justice. But it's also a question of efficiency. And one of the main arguments that's made to justify why it's a question of efficiency comes back to the first point that I started with which is the measurement of labor. Okay, if you look at existing estimates, bring women into the labor force, labor GDP will increase by 6% per year, right? I think this is OECD match. What they do is that they increase L, they estimate a production function with capital and labor and then they basically double the labor force. But that's not quite the case, right? Measured labor might double when you put all women to work but whether labor supply actually increases depends on whether the increase in the market supply is equal smaller or bigger than the fall in the home supply of labor, right? Because if it's identical only measured labor supply increases, right? If not output will increase but welfare will have to take into account the fact that women are working more hours. So when Selim yesterday presented at the G2M League Conference he made the point that when you give childcare women sleep more. So that is a welfare consideration that has to be taken into account. The sleep is very important. I would know that, okay? So even though there might be no change in labor supply once we take into account the non-measurable labor supply there could be a better allocation to buy skills and preferences and the job skills requirement. So there could be a better allocation to market and homework if we allow a gender neutral allocation. And the final thing is whether we can think of an organizational poverty trap. So at the lowest level of development there are few jobs and individual poverty traps. As the economy grows we see organizational change from self-to-wage employment and as jobs become more diverse average earnings increase and inequality falls. So poverty reduction might as a by-product create growth. So we don't have to think about growth creating jobs but other jobs from the bottom. So reducing poverty by creating jobs leading to growth. In other words, we would flip the growth to poverty link. So my last slide is about focusing on talents so aptitudes towards different jobs and how organizations enable people with different talents to work together to attain goals that cannot be attained individually. And this should complement the existing focus on ability and vertical misallocation. And I'll leave you with those questions which is can our poor countries poor because they're poorly organized or are they poorly organized because they're poor? And that's what we'll try to address in future research. Sorry if I went over a bit. We stay up here so you can answer questions. Thank you. Okay, well, Arianna that was wonderful. Thank you so much, Shay. I couldn't write as fast as you were talking. So I look forward to reading more about this and studying it further. Okay, we have 15 minutes for questions. Nikol. Yes, but it's being broadcast. So we should use- Excellent. Thank you. Professor, it was excellent and some things you said like when you commented on unemployment how meaningless it can be. And when you talked about the need for skills and preferences, I liked a lot. And a lot more. But I'd like to focus on two things which drew my attention which I would appreciate your comments. The first is that your three transitions. I was wondering that certainly was true until recently but we now talk very often and there is evidence coming on the more fluid and fluid relationships in the labor market. And I was looking at your framework and it seems like after these jobs within firms there is a sort of fourth transition going on when there is jobs outside firms being created and I wanted you to speculate a bit what would that mean to the jobs variety and everything else you said if we look into those recent developments. My intuition is that everything you said about job variety would still hold. However, may hold in a different setting in which we should bring us back to discussing informality which we try to avoid. And you try to avoid discussing informality by saying that it may depend as you put it by it depends essentially on the registration rules. But here I have a very big difficulty with this because what we see is if this was just registration rules why would it informality not be reduced for 90 years? India was 90% informally 48 is 90 performers. Now surely registration rules could have been changed and the capacity to do them is there. So there's must be something much more deeper which relates to informality and I would like you to comment on that. Thank you again for an excellent lecture. And I have the file back on. So first on informality, I didn't mean to say that informality is not important. I meant that the measures of informality are different in different countries. And so if you put in together micro data at the macro level we do not have a general measure of informality. I think the small self-employment so own account workers is a very good proxy for that. And regarding your second question was actually head ahead 10 more minutes. That's exactly where I was going. So my final slide which you didn't see documented the increase in self-employment which you might call informal self-employment this individual jobs at a low skill which is happening in all high income countries. I think that by now the UK has the same rate of solo self-employed self-employment as Mexico. Okay, so this is happening and it's happening very quickly and these people are completely out of the social welfare net because nobody pays contribution for them. So everything has to be rethought for that. I think that the combination of technology like especially digital technology and this detachment out of firms is gonna create more jobs, more variety of jobs. But at the same time represents a huge challenge for social protection. So I think that is the fourth transition. I completely agree with you. Sure, we'll see. Yes. Can you hear me? There we go. Thank you. That was a brilliant presentation. It was really, really impressive. Thank you very much. Made it worth coming to the conference already. So this is great. But I do have a question about the occupational segregation, right? So first of all, there are macroeconomic effects from the type of gender-based occupational and industrial segregation, right? So when a shock hits a country, it's gonna differentially affect women and men based on that kind of segregation. So, but one of my questions for you then is if occupational segregation increases as the task diversity increases, what would you recommend as the policies to help address that, right? Is it sort of on the supply side where you need to sort of emphasize different skills for people to get more broadly into the lay market or is it on the demand side or both or what would be your recommendation for the combination? So I cannot have my slides back, can I? Can I? Can I have the slides back? Thank you. Because it is an important thing which I didn't have time to highlight, which is that when you compare men and women, actually, if they come back, the women who are, maybe not, well, I can tell you in words, the women are very positive, thank you. That was the self-employment chart. Let me, sorry, I just flashed it very quickly. So if you compare here across countries, men and women, you look at the richest countries, you see that there are very few women in the, so these jobs are East Coast ones. So the higher up, like the orange and the red are the higher East Coast, right? You see that the women are mostly at the top. So there is a very, there is a higher barrier to entry into the labor force for women so that the only women who go in are most qualified and do better jobs. And the jobs are much less varied. So you see there are fewer lines in that graph. So the type of policies that I would recommend, you know, there's a lot of focus on childcare, early years childcare. And I think that's most important for the children than for the mother. I think it's very difficult, it's a very difficult argument to have until mothers stay out, you can't stay at home, you have to go to work. There's a personal choice that I don't think policy should much interfere with. But there are two aspects that policy can change. The first thing I think most important is reskilling for re-entry into the labor force. Because while you can't tell a woman to leave their newborn at home, once children are in school, the mother doesn't need to be at home because the children are in school. So there's much less demand of childcare and there are very educated women who are a complete waste of resource, I mean, completely wasted resource for the economy because these women are educated, they contribute to the labor force. And yet there is no way of re-entering in the labor force after 10 years out. With the current fertility rates and age at first birth, I think that most women would have like 20, 25 years of productive life in front of them and they don't have the option of re-entering the labor force. So once you stop working, you probably stop working for good. So that's the first, I think, and the most important, I think activity rather than focusing on separating mothers from small children, which is not gonna happen. And also it's a very limited, in terms of time, it's a very limited time that we're talking about. And the second thing is, sorry, I forgot what the second thing is, but that was the most important thing. What did I say about subcontracting sleep? That's, well, it might come back to me, but at least I started from the most important that I've forgotten that it would have been worse. And in the back, and then a hand in the front. Well, thank you. I'll be very quick. I'm Jeff Johnson with the ILO Department of Research. I really appreciated your focus looking at what you called the vulnerable forms of employment versus informality. I think that has a lot to do with your time dimension. Did I understand correctly? You were primarily looking at 2000 to 2010. That's the first question in a couple of your slides, because as we have developed, as you know, we have new international standards on the issue of informality, but historical data do not allow those to be done. So I take your point on that and I wonder how that would impact your research. The second point I want your views on the use of education as a proxy for skills. As we know, it's not always the case to where it works well. Finally, when we look now at post pandemic and we look at the changing nature of work and how we balance work and family, how do you think that will affect this data? Because people were working remotely and finding new ways to work. But we have the intergenerational problem now of how people view work and how they move it forward. So I think this is great research and I'd like to follow up with you later, but the issue I see, the nature of work changing very dramatically in the last 10 years. And we always say that every 20 years, the nature of work changes, but we've moved away from the manufacturing sector. And finally, the last point is when you look at these within countries, these differences over time, looking at how these economic structures have changed, the sectoral distributions, I'm wondering how much that will drive some of the differences. So from the bottom, from the last point, we don't do much over time. The only overtime difference that I showed was in the 1800, just to compare the economies at different levels of development for the 1800, data that we have. So all the comparisons are across GDP. They're more or less the same. They're always around 2010 for all the countries. I agree, there are better measures. We should talk about exchanging data. So if you give us the better measure, we'd book very much happily in integrating here. Again, why did we use education? I agree it's not a perfect measure of skill. It's the only one that's available for all those countries measured in a consistent way. And likewise, informality. If you have a good measure of informality and you're happy to share, we're very happy to include it. Thank you. Okay, we have time for one last question in the front. Thank you, Oriana, this fascinating talk. My question is related to what Mika was asking. What I work in Africa, so this is a question. So what you observe in many countries is that you have, let's say, an interrupted transformation. So you have some of this, let's say, urban, particularly capital areas where you do see an increase in wage employment, you see an increase in diversification, but that doesn't have enough traction to bring all the rest of the country in the same direction. So you have this, and then the situation becomes sort of stagnant. You have, let's say 20% of the economy that works in one way and the other 80% that remains basically based on subsistence agriculture or services, et cetera. So, and here in South Africa, actually the situation was even stark, because during the apartheid years, you did have a transformation and a creation of an advanced economy, but the rest of the economy where 80% of the people are operating did not change by design. And that also was not, it's not really, so my question to you is what do you do in these situations? And whether, some of the policies the advisor was talking about, for example, active labor market policies, economic inclusion and livelihood policies, or even public works reshaped to give people some skills, whether you see value in these programs in trying to, let's say, build some traction for this 80% of the economy that is not moving. So it's actually spot on that you say 80%, because that's what we observe in most cases. You either have like between 180% in subsistence informal, or then you just shoot up at the other extreme where you have 80% formal way jobs and 20% not. So there's no economies in between. So that in a way is, as for poverty traps, is both good and bad news, because if you remain stuck there, it's not good news, but the chances are that if you manage to jumpstart and the jump will be very quick. So the issue is how do we get there from where we are? I think, if I think of most programs, it is not really the type of program, but the intensity of it and the level at which he operates. So most programs like sustainable livelihood are at the individual level. Those can unlock individual poverty traps so they can solve the poverty problem of the person that's being targeted. I don't think they solve the problem at a systemic level. So I think we have to think of programs that target entire communities, not individuals within the community. And the second thing is the size of the programs. There is a tendency of spreading resources very thinly. So if you have a certain budget and a hundred people that need help, you divide the budget by a hundred. But if you have these kind of fixed costs that can give rise to poverty traps and increasing returns to scale, giving a tiny bit to everybody is unlikely to have any effect. It might actually be better to give a lot and it's hard to decide who to give it to, but to focus on a sector and push the sector or focus on an area and push that area first rather than try to develop everything simultaneously by giving a little too. So I think that both the size of the intervention and the scale of it that is individual versus community are key. Okay, with that, we're out of time for questions. Please join me in thanking Ariana for a fantastic thought. Thank you so much. Okay, I was given a couple of announcements to make. Please join us for refreshments on ground floor level two that says. So please do. Afterwards, we're going to reconvene here in this auditorium for the plenary Africa policy panel and a question, when do we reconvene? Since we started late, when do you want us back? 1045, okay. So we have 25 minute break now, correct? Okay, again, thank you.