Iterations of a Growth Diagnostic: The Case Study of Albania

Katya Gonzalez-Willette: [00:00:00] Hello and welcome to another episode of the Growth Lab podcast series. The Growth Lab has been engaged in an applied research project with the country of Albania since 2013. In this time we have conducted research on numerous diverse workstreams related to stimulating economic growth in the country. During this research engagement, our team conducted a Growth Diagnostic analysis to understand and test potential binding constraints to economic growth in Albania. After the initial Growth Diagnostic study conducted in 2013, the team has since updated its tests and findings to reflect changes in the Albanian economy over time. In this podcast episode Growth, researchers Jesse Lu and Tim O'Brien discuss the journey through iterations of the Albanian Growth Diagnostic from 2013 to 2020. They delve into the processes and data our team uses to thoroughly study and test various constraints. They also discuss how these constraints have evolved over time and their collaboration with our data development and design team to make the latest Growth Diagnostic update more digitally accessible to a wide audience. [00:01:06][66.1]

Jessie Lu: [00:01:11] At the Growth Lab, we have a really awesome job, we get to work with a bunch of different countries because we work with all these different countries and every country has its own challenges, its own strengths. We at the Growth Lab, produce many different types of analysis, usually responding to the needs of the countries that we work with. Through all these different types of analyzes, one of the most consistent pieces that we produced is something that we call a Growth Diagnostic. And these Growth Diagnostics are kind of the bread and butter of the Growth Lab. We've conducted them for national projects like our projects in Ethiopia and Namibia. We've also conducted Growth Diagnostics at a more granular level. So we've done Growth Diagnostics for some national projects in Loreto Peru and also for city projects like the project that we did with the city of Buenos Aires in Argentina. And so because of the Growth Diagnostics are really important to our work today. We're really excited to share with you a deep dive into the process and analytical approach that we use to conduct these Growth Diagnostic exercises. The Growth Diagnostic is just one thing that of what we do, but hopefully in this conversation we can get a better sense of how we as a growth lab approach problems more broadly as well. And so we're going to use our long-standing project in Albania as a case study. And we're really excited today to share with you what the Growth Diagnostic process is like. So. Hello, welcome. My name is Jessie Lu and I'm a research assistant at the Growth Lab and I've been working on the Albania Project for almost two years now. And today I'm really happy to be joined by my colleague Tim O'Brien. And we're going to discuss the most recent update that we did to the Albanian Growth Diagnostic. This update was led by the two of us and was published in July of last year, which was 2020. And it was a follow-up to a more comprehensive Growth Diagnostic that our team had first published in 2017. And so just to get started right off the bat, Tim, could you tell us a little bit more about what a Growth Diagnostic is and why it's so important for the kind of work that we do here at the Growth Lab? [00:03:05][113.3]

Tim O’Brien: [00:03:06] Hi Jessie sure. Well, I think that the concept of a Growth Diagnostic gets mystified, and to me it's really, really something very simple. It's just whenever we work in a place, we want to understand what the underlying dynamics are, the economic growth in that place. And so the Growth Diagnostic to me is something that's continuous, ongoing. It orients us as a project in our thinking of what we're even trying to do, what we're trying to accomplish in this country or this city or this place, or what problems are we trying to solve, what problems are worth solving? You know, when I took Ricardo Hausmann's class in grad school at the Harvard Kennedy School, I really left after a semester thinking that a Growth Diagnostic was a report. It was a kind of exercise that you do to understand a place and a report that you write that has a lot of components. But now I understand after working with Ricardo for many years in many places, that a Growth Diagnostic is more of a story, a way of thinking and understanding that we have of a place. So in more concrete terms, a Growth Diagnostic really has a few parts. The first thing is you want to ask a growth question or frame a growth problem before you even start, say, running any diagnostic tests. You want to understand what seems to be the problem here is a place not growing or not growing fast enough, or is the way that a place is growing very non-inclusive are there are a lot of people, parts of society left out of the growth of this place, not benefiting enough from it, or maybe the process is unsustainable or at-risk vulnerable. So the first thing that we do in any place is try to wrap our heads around a relevant growth question or growth problem. And then we employ a bunch of tests, diagnostic tests. It's sort of like the scientific method. We pose a hypothesis. Growth in Albania is low because of something, let's say because of the education system, and then we test whether that's true. If that was true, what would we expect to see as signals that this is holding back the growth process? And we look for those signals. We do that with any and every potential constraint that we can think of. So there exists a diagnostic tree, which is a helpful guide in hypothesizing different constraints that could be at play in a place, but we don't hold ourselves to that tree. In fact, when we're engaging with the place, we often get all sorts of theories from people in the government, from people in the private sector, from just people in civil society about what the problem is here. Why is the economy not expanding and or not benefiting as many people as we think it should? And then once we feel like we have a good understanding of that, what holds back the growth process of some kind, then we search for what we call a syndrome. Maybe we found in the diagnostic phase that there's several issues that might be interrelated. We think of a syndrome as a potential unifying story that brings it all together. Maybe there is something even deeper in the nature of the place that gives rise to a number of constraints that bind at different times. And then we keep updating it so we might write a report or we might give a presentation and say, here's our Growth Diagnostic but the world doesn't stop, then it keeps moving. So we want to always keep updating our understanding of the process of growth in the place and how it can work better. So Growth Diagnostic is in some ways never done. We're always trying to diagnose a place better. [00:07:12][245.7]

Jessie Lu: [00:07:14] It is true that the Growth Diagnostic is always an ongoing process, it's never done, as you said, countries are always in flux, the world is always in flux. And I think from my perspective, it's always great to see some areas improving. But even when some things in a country improve, there are always challenges that will remain. And so I think that this theme of always needing to update what we do is really important. And it's one of the strengths of the Growth Lab, especially in really long-term projects like this project we've had in Albania. And so I know that throughout the course of this project, obviously, it started long before either you or I was working on this. But I know that we've done about three diagnostics of the Albanian economy now. And so I was wondering if you could just briefly take us through the takeaways from the first two diagnostics that we did, especially focusing on this really comprehensive analysis that I know you and our colleagues Ljubica Nedelkoska and Ermal Fresheri conducted in 2017. [00:08:05][50.6]

Tim O’Brien: [00:08:07] Sure. So the first diagnostic happened before I joined the growth lab when the project started in 2013, and Albania is in a very different place in 2013 than it was now. So the problem or the question that the team was exploring at the time, the Growth Diagnostic, was why is growth collapsing? Why is the pace of growth collapsing? Why does Albania appear headed toward recession if trends continue? So Albania had been growing, well, I think an average rate of around six percent per year for much of the 2000s up until the global financial crisis and events of 2008. That was after most of the 90s when Albania was an extremely volatile place as the economy more or less grew out of like six decades of closed-off authoritarian rule, it emerged in sort of a chaotic state. But by the 2000s, growth was steady and high. But when the team arrived in 2013 was invited to help policymakers understand economic challenges, better growth was collapsing. So the team wanted to understand why, and it was rather easy to find the causes at that point. Albania was in the midst of a macroeconomic crisis and the macroeconomic crisis had deep ties to the electricity system crisis. So the electricity system was insolvent. You could say people weren't paying their electricity bills and the government was forced to cover the costs of the losses in the electricity system. And the whole fiscal system of Albania was unsustainable. And Albania needed to enter into an IMF program for support to resolve those challenges. But after a few years of really, really successful IMF program and really successful government reform push in several dimensions, in part guided by that initial diagnostic that traced the causes of the macroeconomic crisis to the electricity system and then further diagnosed. Why is the electricity system in such a state of chaos and pretty miraculously in a period of a few years? Those problems were in the past. And by around 2016, or maybe even before it was clear that this was no longer the story of Albania macroeconomic and electricity system slowing down growth. Growth was starting to recover. And we then wanted to ask a different question, which is how can the pace of growth accelerate and the pace of economic convergence with the rest of Europe accelerate. So that led to a bunch of different hypotheses and a bunch of tests for those hypotheses because now we weren't trying to explain growth collapse, but we were trying to explain why wasn't growth faster than Albania was currently seeing. And that was a much more difficult diagnostic. And I was a part of that. And it took a long time for us to wrap our heads around this, because every time we had a hypothesis and we began to test it, we weren't getting strong signals that this problem was binding. So, for instance, corruption is a big problem in Albania. Everybody knows it's a big problem in Albania. So we did our diagnostic test to see if corruption was holding back faster. Growth and relevant diagnostic tests include you want to try to come up with a price or a shadow price of this constraint. What are firms willing to pay to overcome this problem? You can look for changes in the constraint when the constraint is lifted or when the constraint is tightened. Does growth respond? Does private investment respond? You can look for indications that firms are bypassing the constraint. So it isn't such a problem for me. Maybe there's a way I can work around it that's easier for some potential constraints than others like it. When electricity is constraint, firms often have backup power and they use generators, for instance, and a test that we call at the time, Camels and Hippos, which is about the structure of the economy. If there is a very particular binding constraint of one type like water is a constraint in the desert, then you will end up with a pattern where you have a lot of things that are not intensive in that constraint, like camels. And you won't have a lot of things that are intensive in that constraint, like hippos. So we call that Camels and Hippos thinking about water in the desert. But you could think about that kind of dynamic with all sorts of different things. So is the structure of firms very intensive and firms that don't mind if. There's a lot of corruption, are there a lot of firms missing from the mix that you would expect to be deeply impacted by corruption? So anyway, we went through all of the easy to test constraints in the diagnostic tree, access to finance infrastructure like water infrastructure, road infrastructure, electricity infrastructure, the education system. And we were not getting any clear signals across those tests that any of these particular issues were binding. Even when we looked at things like corruption and broader aspects of rule of law, we weren't finding these patterns of Camels and Hippos or clear indications that firms were bypassing that constraint. And we could find distinct periods of time when the constraints seemed to be tighter or weaker and they didn't align with a speeding up or a slowing down in growth. So we were left kind of struggling for a little while to pin down what is holding back Albania from growing much, much faster than it is. And at the time, growth was subdued at around three percent per year. But if there was no meaningful constraint on the process, we should expect it to be much higher and much faster convergence with the rest of Europe around it. So we eventually realized in very interesting ways that this was a natural outcome in Albania, still tied to this closed off past when we started to look very closely at what are the industries present in Albania and where did they come from? How did they grow? How did they start? We could always find or almost always find a first mover, a pioneer, and that often came from abroad. So Greek investment in the cement industry, say, or Albanians who had lived in the US who went back to Albania, helped jumpstart the tourism industry, for instance, or help find new markets for sage and lavender and other spices in Albania, or the garment and textile industry emerged from close interactions with Italy and started through the outsourcing of labor intensive activities in garment production from Italy to Albania, and then grew over time to do more parts of the garment value chain and higher value additions in Albania. So when we looked across all the segments of the Albanian economy, we could trace the origin story of most of them. And we started to see that for growth to accelerate, we needed more origin stories, we needed more industries to appear that could not naturally appear in Albania because nobody in Albania had been exposed to them before. So this dimension of the growth process was slower than the rest. So in other words, there were places in the country that had all the infrastructure that they needed, all of the access to markets that they needed. There were challenges in the rule of law, but not to an extent to hold back somebody from setting up a factory and being super successful. So that became our understanding of the constraint of growth in Albania. And we call that a know how constraint. There's product service, specific knowhow that it takes in order to make a thing, provide a service, especially to export to the rest of the world. And it just has to come from somewhere. You can't create it if you've never been exposed to it before. So that was the process that we thought was most necessary for Albania to accelerate growth. So we started to say, for instance, that the pace of growth in Albania will be constrained by the pace at which it can accumulate Know-How from abroad. And then when we step back to ask, is there some kind of syndrome here? So if that's the problem, why isn't that happening faster? Why is the accumulation of knowhow from abroad and the diversification of the goods and services produced in Albania slow. Then it started to make sense that maybe this is how corruption and rule of law failings start to bind. Because when you look differentially at firms within Albania and potential investors who have maybe approached Albania but decided not to invest, or investors who have never approached Albania but have approached other countries looking to expand, or you look at members of the Albanian diaspora who haven't lived in Albania for a while, we saw a very clear pattern where those who aren't in Albania had perceptions of corruption and other rule of law issues that were much greater than the firms that were there, which was an indication to us that this is a problem that keeps Albania in a sort of low know how equilibrium its image abroad. So some of the actions that Albania would need to accelerate growth would need to overcome that image problem. And also just Albania being so small, many people would never think their business could succeed in Albania. So overcoming that problem as well. So we started to map out our understanding of this low, know how equilibrium. And we started to believe that as things were beginning to look brighter in Albania, it was also the time for more creative policy and creative government capabilities to attract knowhow from abroad. And that drove several aspects of our Albanian project moving forward was to try to develop those potential initiatives together with the government. [00:19:00][653.2]

Jessie Lu: [00:19:03] I think that the story that you really highlight is the story of Albania is a country that really is changing a lot, especially coming out of this kind of closed off past. I think that it's pretty interesting that there was a noticeable rebound in growth after the first diagnostic that we conducted, maybe because there were these kind of big, easily definable things that needed to be fixed. And I think that this thing that you say about the second diagnostic and that process being much harder and also the fact that what we found is that productive Know-How is a huge constraint, but it's not necessarily easy or quick to design policies to help with the accumulation of productive Know-How. And it's very much something that takes a long time. And so given the fact that you identified something that is a little more nebulous, much more widespread, this issue of productive Know-How, and given the fact that maybe it was hard for Albania to grow quickly in response to the fact that we couldn't necessarily relieve this constraint right away, why was it so important that we conduct a Growth Diagnostic so soon after this one in twenty seventeen? Because I know you and I first sat down to discuss this updated Growth Diagnostic in late twenty nineteen. And so what was the kind of motivation behind waiting only two years to update this diagnostic that we had done? [00:20:17][74.2]

Tim O’Brien: [00:20:18] Yeah, well, there were actually a couple of things, I think that all pointed in this direction. So one was, like you said, it's it's sort of a nebulous, hard to describe constraint. It doesn't have easy direct policy recommendations, or at least not as most governments see them. Governments might want to focus on infrastructure or improving the education system or something tangible. But we were coming in and saying, actually, keep doing everything you're doing on all those fronts, but it's not going to be enough to accelerate growth. And that's part of the diagnostic is just having an understanding of what's holding back the process. So sometimes we do a diagnostic and then people end up angry because they're saying, what do you mean we shouldn't invest more in education? And if a diagnostic doesn't point to education as the binding constraint, it doesn't mean to stop investing in education. It just means there's something missing in the mix here. And we should understand what that thing is and figure out if there's a role for government in filling that gap. So this was a new gap for the government. Maybe they had you could think of it as not having had the luxury before to have to worry about this problem. There was always a much more pressing problem. So one thing is that we wrote a big report and we were doing a lot of useful things in collaboration with the government. But the report probably didn't communicate this all that well. Somebody has to read through a lot of text to get to the point and maybe still it isn't so clear. So we wanted to try to innovate more on how we communicated. So first we wanted to check does this constraint still hold? And another motivation for doing that was that. By mid 2019, when we started thinking about updating this Growth Diagnostic in a focused way, growth had kept accelerating to a large degree, at least through twenty eighteen. You know, when you're doing these things, you're always looking back a little in time. So it was in twenty nineteen, but we were looking at growth figures from twenty eighteen and before and the pace of growth had kept accelerating. So we wanted to look back and ask the question of why did growth keep accelerating. That's obviously not a problem. That's good. But did it keep accelerating because the constraint that we had identified was being lifted or for another reason and maybe we didn't quite get it right the first time when we thought it was productive, know how? So we wanted to check on that, check how things were changing. And we also wanted to continue to frame our own project work. We treat this diagnostic seriously and that it determines where we put our efforts. So we were finishing a few areas of work with the government and wanted to know for ourselves where should we use our resources and put our thinking? So for all those reasons, it made sense to revisit the diagnostic, which we did together pretty quickly, I'd say in late twenty nineteen. And just in terms of the growth problem, we found that it had changed with a few more years of data and of course, everything's backward looking. So when we were doing the work in 2016 17, we were looking at growth up to 2015. And by 2019, it had become clear that the growth was really accelerating. And that was exciting. And we wanted to understand why. So the problem in a way was still, can growth accelerate even faster? But it was no longer. Growth is low and lagging and undynamic. It was growth is pretty solid, but it's in line with the rest of the western Balkan region. It's concentrated a bit out of Tirana, the capital and the surrounding area. So we wanted to understand the process better to understand if growth could be more inclusive across the country and if Albania could really become the engine of growth for the whole Western Balkan region and not to trace the growth of the region. And then at the same time, Albania was actually hit by a big earthquake and then subsequent earthquakes, which affected our diagnostic, but only for us to ask how will this earthquake affect the growth process? Is the growth process strong enough to get through this earthquake? Earthquakes knock the process off its path. So for all those reasons, we thought it was useful to dive back in. [00:24:47][268.6]

Jessie Lu: [00:24:49] Yeah, and just to continue off this, Albania face a lot of physical earthquakes and then, of course, we have this metaphorical earthquake of covid, and I think that we can talk about this a little bit more later on the podcast. But I think something that was really interesting for us as we went through this is that when we approached this diagnostic, we had no idea it was going to happen in the future. And we were really curious about what was explaining Albania's current situation. And I think that as we went through this analysis, as covid hit, it really challenged us to reframe what we were writing, to not only be about what explained Albania's growth, but also to understand whether Albania's economic growth and economy as a whole could be resilient against these shocks. And so I think would be cool to dove a little more into that. But before that, I wanted to spend some time talking a bit more about the nitty gritty of the data analysis that we did. So kind of shine a light on this black box of analysis that we do here at the growth lab. So I remember when I first sat down to update this diagnostic, one of the first things that we did is we really wanted to look at the updated available data out there. And so I remember we created a table and we had each row with constraints that we had identified and twenty seventeen or a potential constraint. And then each column was a different survey or data source. I remember we filled in each cell with any new information and then we went through and we rated whether this new information indicated that the constraint had tightened or loosened or whether the constraint had kind of remained the same since we conducted this analysis in twenty seventeen, using, as you said, twenty fifteen and twenty sixteen data. And I remember when we do these kind of analyzes in general at the Growth Lab, we use a lot of general data sources or global data sources like the World Bank Enterprise Survey and doing business indicators. But we also used a lot of Albania specific surveys from the Statistical Institute, which they call INSTAT, as well as from the Bank of Albania. And so I think that this was a really cool exercise because it really forced us to be very systematic and very thorough in ensuring that we were collecting all new information possible in order to update what we knew about each of the constraints. This was a great first step. It was really thorough. It helped us get a broader sense of the big picture. But obviously what we did as that first step was not enough to ensure that we had all the information that we needed to update this analysis. So I was wondering if you could talk a little bit more about the other pieces of analysis that we conducted, the other data sources that we looked at and talk about how this all came together in terms of us piecing a story together. [00:27:14][145.9]

Tim O’Brien: [00:27:17] Yeah, so really, there's no single way to go about this process of understanding how an economy works and what might be broken about it. One thing I like to do is create a big visual table where every hypothesis we could think of as this is the problem that holds back growth or makes growth unsustainable or whatever the problem is as a row and then as columns, the different tests that we can use, shadow price changes versus changes, Camels and Hippos. And then there's things that don't necessarily fall into any of those tests, but just a lot of benchmarking, global reports out there, taking information from those and using it to update our understanding of the place that kind of gets its own column and try to fill in as many of those boxes as possible and start to see what story emerges. So for some of those rows, some possible constraints, we get indication after indication that this is not a binding problem. It's not high price. It's not a change that influences a change in growth, maybe in global benchmarks. Albania doesn't lag behind other countries at its income level and so on. And other things are setting off alarm bells as we go through the tests. They're signaling there's an issue here and then other things will be mixed. There will be some constraints where this hypothesis looks like it could be really important and other tests that say this is not a binding issue. And then if you dig deeper, you might find that the constraint might bind for some sectors of the economy, but not others or some regions of the country, but not others. So that's the way I like to get started. In the case of this, what we call a refresher, because it was not too long after the previous one, we laid out all of our conclusions for each constraint that we looked at and then started asking the question of has this constraint moved at all? And that's what you're describing. And the first very easy thing in my mind to do is look at standard reports that come out frequently, like the World Bank doing business indicators or the Global Competitiveness Index, all sorts of reports like that. And just look for hints. Those reports have a lot of detail in them. And sometimes if you dig really deep into the tables, you find signals that that something is weird here. And then that's a starting point for me, is then to ask the question of this weird thing. Can I apply one of the other tests on it? We also often use the World Bank's enterprise survey a lot because the World Bank's enterprise survey is conducted all around the world every few years in each country, and they speak to actual firms, whereas the doing business indicators speak to people who understand the laws. And when you speak to the actual firms, you understand how they actually act and they might not actually act in accordance with the laws. So the enterprise surveys are really valuable because you get very clear signals of if firms think something or another thing is a constraint and then you can dive deeper and deeper. So there's really no roadmap. We would like to start with those sources and then ask further questions. And sometimes the way to get information is to then go to government offices themselves. For instance, the investment promotion agency AIDA, in the case of Albania, speaks with potential investors. So they know the landscape, they know what investors are saying. And that can be very useful in in diving in and actually finding that maybe the constraints to investment in agriculture are different than the constraints to investment in manufacturing, say. But we end up using very different data for each different constraint. So when looking at electricity as a constraint, we're looking at using national electricity consumption data and supply data for transportation, which you are a part of. We actually used Google Maps to look at the changes in travel time from when we documented them in two thousand sixteen seventeen. Then we used Google again in twenty nineteen and checked how travel distances changed and how travel times changed. And those are very clear indications of how the transportation network changed. And we actually found that for some parts of the country, transportation connectivity improved and for some parts of the country it worsened as main arteries got. More congested. For access to finance, we looked at Bank of Albania data on the banking system for use on the tax system, we looked at the rates themselves, but also local surveys of firms like the American Business Chamber in Albania speaks to firms regularly and has annual reports that go into depth on their views on things like the tax system and beyond. So really a whole bunch of different data sources and we dig them up. We're always finding new data sources based on the need. So when we look at that big table and we see this row, looks like it might be a constraint, but I don't have any information to fill in this test. Then we go on a search for is there information out there? Can we think of a new way to run this test? There are other international sources of data that came out between the two years, like PISA scores for education quality when testing education as a constraint. We also always look at returns to education. So for that, we use labor for surveys in Albania and econometric ways of looking at returns to education, returns to secondary education versus tertiary education and so on. We're oriented around this. What are the sources of foreign know how? We are also able to look at our own surveys of the diaspora and understand how diaspora views are changing. We used the FDI markets data set as a data set from the Financial Times that registers every internationally announced investment project. So we were able to look at the nature of foreign investment into Albania in comparison to other countries that we had hoped that Albania would follow their path and see to what extent Albania is attracting foreign investment in the ways that, say, Romania did in the mid 2000s or the ways that North Macedonia did more recently or further away in the ways that Vietnam has and other countries that are that can be thought of as something like role models for Albania. [00:34:31][433.8]

Jessie Lu: [00:34:33] Thanks so much, Tim, I think that's a really good overview of all the different unique and diverse sources that we really draw from as a lab in order to get a better sense of what's happening in these countries. We supplement a lot of the stuff by traveling to countries. But with everything going on, I think it's important to know also the strength of the sources that we have that do not necessarily require us to go to countries physically. I would just add that another interesting thing when we were going through this diagnostic was this constraint that we had identified called rule of law. And I think that this was a really interesting example in the sense that we kept getting different signals. I remember in the initial pass that we did, it seemed that rule of law was getting worse in Albania and it seemed like this was becoming more binding and it was going to be a bigger obstacle for the country. But I remember we started piecing together a bunch of other sources. We looked at global reports on rule of law. We also looked at news updates and what the media was saying in Albania. And I think through all these sources, we found that rule of law was a big issue, but it wasn't going to be binding. And I think that that exercise was really interesting in terms of highlighting the fact that oftentimes signals come from all over the place and it's all about really being thorough and finding the information that we need to find and always asking how it fits into our broader story, whether it supports what we're saying, whether it forces us to reexamine what we're saying. And I think that rule of law was an interesting example that came out of this refresher process that we did. I just really want to focus in now. We've talked about all of these kind of other constraints that aren't as binding, things like electricity, road transportation, rule of law. But I want to focus a little bit more on this productive knowhow, which you said was the big constraint that you identified in twenty seventeen. Could you talk a little bit about what we found in terms of how productive Know-How had evolved when we dove back in in twenty nineteen? [00:36:22][109.1]

Tim O’Brien: [00:36:24] Yeah, so it was very interesting. And exciting because we found that growth seemed to be accelerating precisely because the productive know how constraint was moving, but it was being lifted rather gradually and slowly, which is necessary when you think about this kind of dimension of, know how moves slowly across borders. But it led us to ultimately in our project, double down on approaches that might accelerate this process. So there are a lot of distinct measures that we were able to trace that showed this improvement in know how. For instance, I was speaking about the World Bank enterprise surveys before the World Bank Enterprise Survey for Albania was done in 2019. The last one and the previous one was in 2013, and there was one before that in 2007. So we were able to look at changes over time, as well as comparisons against other countries. When we looked at changes over time, we found almost universal indications of increasing know how. So the age of firms was increasing the average age of firms, the average years of experience of managers who were surveyed, the proportion of firms that offered training, the proportion of firms who said that they had introduced an innovation in their products or processes. Over the last year, there was an increase in the tendency of firms to export. There were also indications in the enterprise survey of a tightening labor market constraint. That is, firms reporting more often that they were struggling to find qualified workers. So these are all signs just from that survey alone that the Know-How of those firms was increasing. We could also see from administrative data that more and more people were working in larger firms, which is a process that you would expect to see as the private sector becomes more sophisticated and also that the average wages in those larger firms were driving up wages overall. So this was a really healthy process. But in our own indicators of economic complexity, on our atlas of economic complexity, we saw only gradual movement upwards in Albania's economic complexity index. So it was improving, but rather slowly. And on the Atlas, we also project what a country's growth would be over the next 10 years, based only on its export complexity in comparison to its current level of income. And that projection for Albania at the time was three point three percent. Albania was growing at four percent. So it was the ECI of Albania had not yet surpassed its potential. So this dimension was gradual when we zoomed in on different sectors of the economy and ask the question of for exports in particular, what has been driving export growth. We found that it came primarily from tourism from IT. And when we looked at the businesses operating in those industries, we could see all sorts of dynamism of new types of businesses being introduced, often coming from abroad. When we looked deep into agriculture, we could see that the number of products, the diversity of agriculture products was increasing. The number of markets that each agricultural product was reaching was increasing. There had been an emergence of different types of fishing and aquaculture that had been constrained for a long time. And we could trace that emergence and rapid growth to the entry of foreign investment from places like Spain and Italy. So Spanish investors producing aquaculture products and then exporting them to Spain. So we could tell that there was movement here, but we could also see that there was space for faster movement. And I was mentioning the FDI markets data before we looked at all of the industries in which Albania had had at least one foreign investment project announced over the last few years. So, for instance, there had been two an auto parts. There had been some in fishing, like I mentioned before, and we compared to a set of other countries in the Balkans and beyond. And we could see that the number of new investment projects that Albania was getting was super low and undiversified in comparison to these other places. So all this indicated that the process was happening, but that it could happen much faster and that growth could be much quicker. And on the dimension of foreign direct investment, it could probably be more inclusive across the country if investors knew not only that they had an opportunity in Albania, but that they had an opportunity in the south of Albania or the north of Albania and could more easily find it. So it was, in a way, a confirmation of the previous constraint, but also an indication that the constraints still held. [00:41:35][310.9]

Jessie Lu: [00:41:37] And I think it will be interesting for the next diagnostic that we do and who knows when that will be to continue to monitor how this changes and see whether Albania can converge with what we predict in the atlas and to see what happens with this economic complexity. I just want to take a second and think about the fact that when we we first constructed one draft of the story before we all went home in March of 2020 last year. And I think that once we really sat down to figure out how to publish this, we had to reframe the story a little bit in response to what was going on with covid. And so I wonder if you could just really briefly talk about how we approached this reframing and what we found when we took what we had done and then kind of changed to the question. So the question was not only what is going on with Albania's growth, but the question became, as we titled, our refresher, can Albania's economic turnaround survive this covid shock? [00:42:33][55.3]

Tim O’Brien: [00:42:34] Well, yeah, it's a tough question, looking back at that time, Ricardo calls the first few months of covid-19 the fog of war. When we were asking this question, we didn't even know. Nobody knew how long this crisis would last, how it would evolve. So even when we wrote and published this brief discussion with the title, Can Albania's Growth, is it resilient to covid-19? We didn't even know what covid-19 was, really. So in my mind, I expected it to be a shorter, lasting constraint than what it turned out to be. But it pointed to vulnerabilities. For one thing, since we could see that the acceleration had so much to do with one, tourism was scary for Albania because it was in the middle of an enormous tourism shock and at the time was asking the question, will there be any tourist season in the summer in Albania? But on the flip side, the other really interesting driver was information technology, business process outsourcing, a knowledge process outsourcing. And the pandemic actually accelerated the global emergence of some of those types of work. And Albania is positioned to benefit from that. I also spoke about agriculture. There was an initial period where Albania, like everywhere else, suffered enormous supply disruptions in agriculture, but like many other places, also quickly rebounded. And because of the lack of access to some markets all of the sudden or lack of access to some inputs, all of the sudden the sector transformed or continued its previous transformation, which was good. So it's still actually an open question is how will this process recover once global travel recovers and once global vaccination recovers? My sense is that there is a very strong foundation here in Albania, but the shocks of the pandemic were large enough to introduce real macroeconomic crises, of course, not just in Albania, but all over the world that need to be dealt with. But the fundamentals of the diversification and increasing complexity of the place are solid. And we see this in some ways as an opportunity for Albania to leapfrog in its abilities to attract diversified investment, knowing that the global economy is in the midst of reshaping itself to be more resilient in light of how everybody was affected by the pandemic. So the world changed. The diagnostic may have been accurate in twenty nineteen, but the world changed a lot in twenty twenty. So if we were just to write a report on what constrains growth in twenty twenty, it would be covid-19. But the exercise pointed to the post covid-19 future as well. And we think that in the aftermath of covid-19 growth will once again be constrained by how fast Albania can accumulate Know-How from abroad and that the pandemic itself introduces new opportunities. Don't let a good crisis go to waste to activate initiatives and active roles of government in order to plan and prepare for that post covid future. [00:46:07][212.7]

Jessie Lu: [00:46:09] It's a very optimistic outlook, and so I think we'll just have to wait because only time will tell. I like experts on what's going on with COVID in Albania. We work on a very large team. And so I should say that if you, our listeners are interested in learning more about the details of what happened with covid in Albania, you can listen to our first team podcast in which our colleagues, Shreyas Gadgin Matha, Ricardo Villasmil, and Ermal Fresheri kind of go through the health and economic impacts of the pandemic in the country. I just want to end and spend a little bit of time talking about the format of this refresher that we did. So you said at the top of the podcast that this refresher was really interesting because we were faced with the challenge of trying to translate a really long, wordy, analytical diagnostic, which was the version that we had published in Twenty Seventeen. We wanted to try and kind of convert this into something that was a little more digestible, easily accessible. And we had thought initially about writing a brief, but we were encouraged by our good coworker Shreyas Gadgin Matha to kind of think a little bit more outside the box about what we could do with this. And so this diagnostic was really unique in the sense that we partnered with our Web development team where the dream team there behind our beautiful tools like the Atlas of Economic Complexity. And we partnered with this dream team because we wanted to create a more interactive piece with dynamic visualization. So why was this Growth Diagnostic refreshers such a good opportunity for us to try this new format? Who are we trying to reach? And also, what was it like working with the Web development team? [00:47:41][92.3]

Tim O’Brien: [00:47:43] Well, working with the development team was spectacular. It created a lot of new ideas in terms of how to visualize a graph and tell a story with a graph that I had never known how to do before. It was also great in their focus on product design. So often when I write, I just start writing. But with the design team, it's a very conscious process of working backwards and setting timelines and building in space for iteration. And that was a really wonderful opportunity for us in this project that I think we're going to use in other projects as well. It was a good opportunity because, as you said, this was a refresher. We had a long document. You know, even in this refresher, we have an internal PowerPoint presentation that's over one hundred slides and each slide is a piece of evidence. And to us, each slide is interesting in its own way. But we're conscious that the world doesn't have time for that, Albanian policymakers don't have time for that either. They want us to get to the point, bring things together in a way that a broad audience can understand. You don't have to be an economist and you don't have to be an economist who's happened to use the framework of Growth Diagnostic before. So the fact this came on the backs of a longer report was a nice opportunity to challenge ourselves, to be brief, engaging in the way we presented graphs and evidence and tell a story. And that, I think turned out really well. It forced us to simplify things, to make cuts, to ask ourselves what have we learned that's really relevant. So the audience was our normal audience of policymakers in Albania, policymakers and development professionals worldwide who are working in all different types of contexts to understand growth processes and different types of places and figure out what are relevant policy actions to address constraints, but also just larger segments of Albanian society in Albania and Albanian diaspora. We were hoping would find this refresher on the VISHUB interesting. [00:50:09][145.9]

Jessie Lu: [00:50:11] And we did end up sharing this a lot with the policy makers that we work with in Albania and with our networks more widely, and I think that in the process of doing that, that was also a really interesting opportunity for us to see if we got the story right and to see what other people said in response to what we did. Is there anything else, Tim, that you want to add about this before I close out the session? [00:50:32][20.6]

Tim O’Brien: [00:50:34] Well, Jesse, you've worked at the Growth Lab for a few years now, and the concept of a Growth Diagnostic was foreign to you, you came to the team after a long project in Albania. So I would just love to hear your reflections on when you had come from outside the Growth Lab. Is this what you imagined working with a country would be like working on a Growth Diagnostic would be like? And what are some of the things that surprised you? What are some of the things that you think will be skills that you take forward for the rest of your career? [00:51:13][38.8]

Jessie Lu: [00:51:15] Yeah, it's a really interesting question. When I first joined the Growth Lab, I didn't know the Growth Diagnostic was. And I think that what was really valuable to me about going through this refresher experience with you when I first started at the growth lab is the idea that it's possible to systematically assess what's going on in a country. I think pure objectivity is very difficult, especially as development practitioners. It's always hard to know if we're on the right track or not. But I think that what's really interesting about the process of conducting a Growth Diagnostic is that it really forces us to be systematic about what we're doing and to go through the tests and to go through possible constraints and to go down the diagnostic tree. And I think that that is really interesting and really unique about what we do at the Growth Lab. And I think that since working on this refresher with you, Tim, I've also worked on much more comprehensive Growth Diagnostic in Loreto Peru and in Namibia. And it's been really interesting to think about how the Albania project can kind of take these basics that the Growth Lab does and take them a step further when we have such a long engagement with the country. So the Peru project was really unique. It was only six months, this Namibia project, we're right at the beginning of a three year engagement and the Growth Diagnostic that we've done there are very, very thorough. They go through all the tests, they go through all the possible constraints. And we're really at the beginnings of trying to construct a story. But I think that the Albania project is fascinating in the sense that we could pretty quickly do this refresher. And we had a good sense, I think, of how this refresher built on the story that we had constructed as a team and our predecessors have constructed as part of the Growth Lab. And I think that this story is really fascinating because it continues to define the work that we do as a team with the Albanian government. We don't necessarily work directly in Growth Diagnostic or economic complexity with Albania anymore. We work in more advanced topics like how to improve the electricity market or how to engage the diaspora to leverage the know how that they've gotten abroad. And I think that this entire process of joining the Growth Lab going through the Growth Lab, but also working on this Growth Diagnostic refresher and then working on a lot of other things since we've finished this refresher has really highlighted the power of long term engagement, the power of what we do at the Growth Lab. And it's been really an honor and a joy to be a part of this analytical process. So to close off, I'll just say all of the findings that Tim and I found are all available on the website and on this Vizhub that the Growth Lab now has to house all of our really fantastic visual tools that are coming out of our Web development team. And thank you so much, Tim, for sitting down. I don't know if you have anything else you want to say. [00:54:03][168.1]

Tim O’Brien: [00:54:04] Thank you. It's been a pleasure to not only talk through these various Growth Diagnostic in Albania, but also to reminisce about that fog of war that we spent together on this latest update. Thank you for the opportunity. [00:54:22][17.9]

Katya Gonzalez-Willette: [00:54:25] If you want to learn more about the Growth Lab's latest research and events, please visit growthlab.cid.harvard.edu. [00:54:25][0.0]