The Science of Everything Podcast - Episode 103: Introduction to Economic Growth and Global Poverty

Episode Date: March 31, 2020

In this the first in an eight-part series covering the causes of economic growth and development, I provide an introduction to the key concepts of absolute and relative poverty, different theories of development, and how GDP is defined and measured. I also discuss the differences between rural and urban poverty, give an overview of the different levels of development of various countries around the world, and outline some of the major methodological challenges in studying causes of economic growth around the world. Recommended pre-listening is Episode 12: The Price System, and Episode 56: The Gains from Trade. If you enjoyed the podcast please consider supporting the show by making a paypal donation or becoming a patreon supporter. https://www.patreon.com/jamesfodor https://www.paypal.me/ScienceofEverything

you're listening to The Science of Everything podcast episode 103, Economic Growth and Development, Part 1, Introduction to Poverty and Growth. I'm your host, James Fodor. So, at long last, after many promises, I'm finally able to begin my series of economic growth podcast episodes. As I said, this is going to be, if all goes to plan, an eight-part series,

where I'm going to try to endeavor to explain what is known about basically the reasons why some countries are rich and some countries are poor, with a focus on trying to explain the divergent growth outcomes in different countries over the past couple of centuries. So some countries have performed very well, some countries have performed moderately well, some countries have performed very poorly. The purpose of this series is to investigate the reasons and the research that have been conducted into why this is the case.

So I've done a lot of research preparatory for this point, episode, including consulting literally hundreds of papers and doing some of my own sort of statistical analysis and looking at data and reading a number of books. And this is something that I've been working on, kind of on and off, obviously, for about two years now. So I hope that you enjoy it and find it interesting. This is obviously a very important topic in terms of understanding, you know, many aspects of the current kind of geopolitically economic state of the world. And in the course of this series, we'll talk about a whole range of issues relevant to that in terms of politics and economics and other issues. So in the first episode, I'm going to focus on an

introduction to basic concepts of poverty growth, how we measure and define these concepts and have a look at different regions of the world and an overview of, you know, kind of what countries are where. In the second episode, I'm going to focus on the kind of when and where of growth. So this is a history of the world economy. I'm going to talk about what pre-modern economy is are like before the Industrial Revolution and talk about the Malthusian Trap. I'll talk about the concept of the Great Divergence, where sort of Europe began to diverge from the rest of the world in the early modern period. Then I'll talk about the first industrial revolution in Britain and the second industrial

revolution in continental Europe and the US, the Great Depression and the area of central planning, and the Breton Woods era in the decade succeeding World War II, and then the modern period of near liberalism and globalization. and so we'll cover all of the issues of the history and details of that. So you can have a picture of sort of the historical landscape. In the third episode where we focus kind of on the what of growth, that is like what happens when a country undergoes economic development. The focus of this episode, Part 3, will be on the structural changes that occur in an economy when it grows.

So we'll talk about Rosto's stages of growth and changes in agriculture, urbanization, capital accumulation, that goes along with that. So that's not really an explanation of why growth happens, but just sort of what it looks like when a country experiences growth. In part four, we're going to talk about the sort of how side of growth, which is focusing on economic growth models,

which are generally formal mathematical models. Obviously, I can't go through them in detail in an audio podcast, but I can talk about their main features and talk about how the models have become more sophisticated and focusing on different things over the decades. So beginning from the Harrod-Dimar model in the 40s through the Solo Swan model, Ramsey model,

Romers, Spillover's model, Indogenous Growth Theory, and Coordination of Failure Model. Of course, don't worry if you don't know what any of those things are. We will talk about it when we get to it. Then finally, in episode five, we get to sort of the real core of the issue, which is the why side of it. Why do some countries experience growth, and other countries don't?

And so, in part five, we're going to focus on what I call single-factor explanations, which are promoted by various scholars, basically focusing on a single type of explanation. So we're going to look at geography, culture, dependency theories, democracy, education, and institutions as sort of single factors that various scholars think are sort of the key to growth. And the focus of the episodes after that will kind of be mostly on the institution side of things, which I think is the most promising of the single factors, although obviously I'll talk a lot more about that in episode five. But then so in episode six, we'll talk about how growth begins. in low-income countries and why it's able to start in some very poor countries and not others. Again, focusing mostly on institutions and policies.

In episode seven, we'll talk about how growth is sustained in middle-income countries, so they're sort of going from low to middle-income and middle-to-high-income, and those involve somewhat different processes, and so in episode seven, we'll talk about that. And then in the final episode eight, I'll talk about some key sort of policy debates, like trade liberalization, capital and labor market liberalization, privatization, and the theory and evidence relating to the... those and their effects on growth and then conclude with some summary and overall reflections on the

series. Final thing I should say is in terms of recommended pre-listening, it's not really that essential, although if you've listened to some of my previous kind of politics-economics episodes, that would be helpful. A useful episode would be episode 56, the gains from trade. That's quite relevant for some of the later episodes. And also, episode 12 on the price system as a kind of basic introduction to how prices work will be relevant for understanding some of the things that I'm talking about here. But I wouldn't say those are essential, just kind of useful background listening. So, that all being said and that introduction out of the way, let's jump into it and start talking about poverty and growth. So again, the focus of this episode is kind of just on the

background concepts, what we're talking about here. So let's start with the concept of extreme or absolute poverty. Now, this is defined by the United Nations. as being a condition characterized by a severe deprivation of basic human needs. So food, safe drinking water, sanitation facilities, health, shelter, education, and so forth. It depends on income, but also on access to other services. So the basic idea and difference between extreme poverty and what we might just call poverty is that poverty can refer to a sort of a threshold or a state of deprivation relative to some socially accepted standard.

So there is a concept generally used in wealthier countries called relative poverty, although it may just be called poverty for short, which defines poverty as below a threshold relative to everyone else in the country. So a common definition is you're in poverty. If you earn less than 60% of kind of the median income, which is sort of putting you in the middle of the income distribution, so 6% of that puts you in poverty. Now the thing about that definition is it depends on whatever else is earning.

So as the country gets richer, the poverty, poverty line goes up. But absolute or extreme poverty isn't defined like that. It's defined in terms of kind of an absolute basic human needs. And obviously that's a little bit subjective to define. But the current definition places the international poverty line at $1.90 per day in 2011 prices, which in 2018 prices is about $2.12. Now, I should say, I think I'll discuss this a bit more later, but when I say a dollar in really any of these cases, these are what are called international dollars. that the differences in prices between different countries have already been facted into this. So we'll talk a bit more about this later, but because obviously different countries use different currencies

and goods cost different amounts, that needs to be facted into when you're making comparisons. And so this use of international dollars, which are often just expressed in terms of equivalent US dollars to buy a similar basket of goods. So whenever I talk about dollars really at all in this series, that's what I mean, that this sort of international, what's called purchasing power parity adjusted. dollars. So let's call it about $2 a day. You may have heard a dollar a day. That's kind of a slightly older amount, adjusted of inflation. We're talking about $2. So that's the baseline of what's considered to be absolute or extreme poverty. That might seem extremely low. It's very hard to

imagine living on $2 or less a day in any developed country. A big reason for that is simply that it's generally illegal to sell goods and services that are of the sorts of quality that would be purchased by people who in the developing countries are actually living on that amount. So in terms of like accommodation, it's not really legal, I think, pretty much in any developed country to, you know, rent out accommodation that doesn't have running water or electricity or meets basic safety standards and so forth. But actually a lot of people in who are in extreme poverty do live in accommodation of those, of that sort of level. Likewise, you know, in terms of food quality and other things like that, it's much lower than anything you'd likely be able to even be allowed to sell or, that anyone would want to buy in the developed world. So I just emphasize that because it's important not to think in terms of draw too much

from our own experience, assuming most of my listeners are from developed countries, in terms of, well, how could anyone live on $2 a day? In fact, a lot of people do in the developing world, but you can't make a direct comparison to developed countries. Extreme poverty does not exist in developed countries. There are a few people that have claimed otherwise, but without getting into the details of that. The way extreme poverty is defined and measured, it's basically not existed in any developed country. And we'll talk about what the developed countries are in a little

bit, but Europe, US, Australia, where I'm from. Essentially, no one is living in extreme poverty in these circumstances. There are a lot of people who live in relative poverty, but again, there's a very different, that's a very different baseline. Relative poverty is still a big problem, but the level of deprivation is much, much higher for extreme poverty. And so that's a natural point of focus in terms of the people most in need. Poverty is best measured by consumption rather than income. So, of course, in any country, there's people who don't earn income or even in negative income if they run a business and make a loss. But that's not really a very good measure for poverty, because it's really the resources you have access to for

consumption purposes. And in particular, even people who don't earn any income still have generally much higher consumption levels because of personal savings, work that they do that's not reported in official income statements, benefits they get from the government, private charity, use of public facilities and services, help with family and friends, or just living in a household where they get access to household services. So these will provide additional consumption possibilities. And particularly those sorts of non-income-based consumption possibilities are generally much, much larger in developed countries and they are in developing countries. So for example, even homeless people in, at least most developed countries, have access to public facilities like, you know,

bathrooms and libraries and, you know, safe drinking water that's available there and soup kitchens and other things like that, and often internet access and other things like that, which serves as a lot of people in developing countries don't have access to it all. So the important thing is just to understand the differences in these concepts and how they are quite distinct. So about 200 years ago, about 80% of the world's population was thought to live in absolute poverty, and many of those in the 20% who weren't that far above it. Today the number is about 10%.

So that's about 700 million people. Most of those in the 700 million live either in India, China or in sub-Saharan Africa. And there's a smattering in Latin America and other places, as we'll talk about in a moment. This is a common misconception, or I don't know how common it is, but it's at least somewhat of a misconception, that most people living in extreme poverty are not starving to death. They get generally enough calories to live. Actual starvation is extremely rare outside of very specific famine incidences. So two studies from Africa indicated that only about 2% of children were experiencing actual starvation,

in some of the most impoverished countries. Now, many more of those children would be experiencing malnutrition, but that's not the same thing as starvation. Starvation is really a significant impairment of the amount of calories that you're able to take in relative to basic metabolic needs. Malnutrition is an inadequacy of the diet to provide all of the nutritional requirements, vitamins and minerals and fiber and other things like that.

So about 10 to 15% of the world's population experienced some form of malnutrition. So that's many people who are experiencing extreme poverty, experience malnutrition, but not the same thing as starvation. So, you know, there's this sort of idea in the rich world about, you know, starving children in Africa, and there's those images from, particularly the Ethiopian famine from the 80s of, you know, these starving children with extremely thin limbs and the big bellies. That's a very rare condition.

It does happen. Mostly it's in acute crises in some of the most impoverished countries. but just be aware that that's not really what we're talking about when we're talking about extreme poverty. And I think that that's an important misconception to understand because if you conceive of extreme poverty is basically, well, people just literally don't have enough to eat, then the answer seems pretty straightforward. Well, you know, they just need food, right? But usually people in extreme poverty don't really just need food aid other than in, you know, acute crisis situations like after a natural disaster or a civil war. But in most cases, that's not really the problem. And as we'll talk about, it's much more complicated than that.

Okay, so that's some basic ideas of development and poverty and what we're talking about. Let's talk a bit more about conceptions of development specifically. So there's poverty which we've discussed and then there's development. Development is a much more contested concept, and there are many theories of what it means for a country to be developed or underdeveloped. So the sort of classical modernization theory of development says that basically there's a transition, a progression from pre-modern or traditional countries and economies to modern, industrialized, developed countries. This idea originates from the sociologist Max Weber, and it emphasizes the importance of social and political changes, increasing focus of rationality, democracy, growth of industry, urbanization, applying science and technology, and building up the sophistication of the economy.

And so basically modernization theory regards development as equivalent to economic growth and modernization. Though not often expressed explicitly in that term, modernization theory basically forms the basis. of much modern economic analysis in terms of the causes of growth, which is mostly what I'm going to be focusing on in this series of podcast. So implicitly, I'm kind of coming at the question from a modernization theory, although there are aspects of the human development, which I'll talk about in a moment. But I wanted to emphasize that there are other approaches to development as well, and so I'll just briefly talk about those. So there's the concept of human development, which is a more recent idea. Instead of focusing on the sort of society and the economy and the level

of industry and technology that's used, human development focuses on the human condition and the things that people have access to that enable them to live the sorts of lives that they want to. So there's a strong focus on health education and gender equality, for example, as well as human rights. And it's related to what's called the capabilities approach pioneered by Amar Senn. And that focuses on basically the things that humans are able to do and the barriers that prevent them from exercising their capabilities. There's a lot more to say about that and it also has philosophical aspects that I don't really want to get into here. But the idea basically is that human development has a more human focus, particularly health

education, human rights. And I will talk a little bit about that, but it's not my primary focus here, because basically the sorts of questions you might ask and the evidence you might look for in terms of trying to understand human development is mostly about programs to improve health education and human rights. And those are rather different, at least in my view, and I think in the literature bears this out, to the sorts of questions you ask about why do some countries experience more rapid economic growth? And it's those latter questions that I'm more interested in. And I'll talk about why I think that those are more interesting later on. But so, that's why I'm not focusing as much on the human development side of things. There's also an even more

recent concept called sustainable development, and this is probably one of the more popular ones today, particularly because we have the sustainable development goals from the UN. The notion here is that sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs. So there's often a strong focus on negative environmental effects of economic growth and industrialisation, such as pollution, by diversity loss, resource depletion, and especially more recently, climate change. There's also focus on the social institutions and cultural aspects of growth, so the idea is is you need development to be consistent with maintaining robust and functioning social institutions

and ensure the development benefits everyone. So there's also a focus on equality as well. And so one way that this is expressed is in terms of the three different spheres, environment, economy and society, and that they kind of all need to be working together in order to have sustainable development. As I said, this is a fairly popular concept today

for people who study sort of development specifically. My approach is more coming from the economics literature and economic growth, which is a little bit different. Not that I don't think sustainable development questions are important, but just I think that they are somewhat different from the type of questions that I'm emphasizing. I also think that many nations don't have the social or political desire to engage in a lot of sort of environmentalist policies until they reach a certain level of development,

and they can kind of afford to do that. If people can't even meet their basic needs, often the idea of resource depletion or climate change is kind of the least of their worries. So there's more to say about that, but that's one of the reasons why I'm trying to focus more on the question of growth. One final thing that I'll mention, that theory is called post-development theory. This is a little bit more, I don't want to say fringe, but a bit less mainstream. But I just wanted to mention it.

The idea here is basically a rejection or a very substantial restructuring of the way that development is understood. Post-development theory argues that the whole concept and practice of development is kind of a reflection of Western colonialist hegemony, over the rest of the world. So it sets up a hierarchy, according to this theory, of the developed and the underdeveloped nations, and the developed nations are seen as more advanced and superior, and the underdeveloped nations are thought of

as needing help and wanting to be like developed nations. Post-development school argues that such views are Eurocentric, they're kind of universalists, so they're saying that there's the same kind of social and economic progression everywhere, which they often reject. They think that they're based on Western models of industrialization, they're unsustainable. There's aspects of the sustainable development

critique in here as well. And they think that these kind of models are often ineffective because they don't consider the local cultural and historical contexts of the people that the policies are applied to. Post-development theorists generally want to promote pluralism in development. So what development looks like is not the same in every country. It's not always going to be, you know, fossil fuel-led industrialization and building roads and all this other stuff. It might look quite different. It's not really quite clear what post-development theorists think it should look like. I think part of that is because it pretty much explicitly rejects any universalist answer to that. I don't think too much of post-development theory, partly because basically I do buy into the modernization theory idea that there is a sort of a meaningful objective difference between an undeveloped and a developed country, and that you can make reasonably universal statements about that.

And we'll discuss more of that as we go forward. But I also do think that basically people in all countries want their countries to become developed. because it allows people to lead longer lives, be more educated, fulfill their needs better, and have more economic security, and just enable them to do more things that they would like to do. So that's the basic focus about why I think growth is particularly important and why the focus more or less will be on modernization theory, while acknowledging that there are different concepts of development as well. Okay, so at this stage we've talked about extreme poverty, and I've talked about development, The idea being that if a country is underdeveloped, it has a relatively traditional economy,

which is relatively not industrialized, and produces low output per person, and that leads to extreme poverty. So this leads us into the next little topic, which is how we measure economic activity, and this is very important for discussions that go ahead. So there's a concept called gross domestic product. I can't remember if I've talked about this before on the show, but we're going through it here again, if so. gross domestic product is abbreviated GDP, so I'll talk about that a lot throughout their series of episodes. GDP is gross domestic product. It's a monetary measure of the total market value of all the goods and services that are produced in a particular time period in a particular country. So generally it's a year.

Within each country, GDP is normally measured by a national government statistical body and normally uses information that private sector organizations don't have access to. So there is a question about the validity of that data, and there have been some concerns raised about this, especially recently with respect to GDP data from sub-Saharan African countries often don't have very well-funded or competent national statistical bodies, and also with respect to China, which has been thought to be cooking the books in various ways to make their growth look better than it has been. I'll talk more about that later. So Robert F. Kennedy famously criticized GDP as being a measure of, quote, everything except

that which makes life worthwhile. He also said that, quote, it does not include the beauty of our poetry or the strength of our marriages, the intelligence of our public debate, or the integrity of our public officials." Now, that's all true, and GDP has come under quite a lot of attack recently from, I mean, that's an old quote, but the attacks on it continue. Basically, the feeling is that it's overused and that it doesn't really reflect how good life is. It's a measure of the productivity or the production in an economy. The reason I think GDP is important, and many economists, well, most economists really still use it as kind of a primary measure of development and economic output, is that many of the things that we do think make life worthwhile, maybe

not the strength of our marriages so much, but other things like beauty of our poetry, intelligence of our public debate, and integrity of our public officials, and many other things like health, education, transportation, communications, leisure, access to information, and many more do depend on the productivity of the economy. So it's hard to have an intelligent public debate if nearly everyone is illiterate and they don't have access to mon communications and it's difficult for people to travel from one part of the country to another to attend events or other things like that. It's difficult to have good poetry if, again, hardly anyone can read or has access to that poetry and almost everyone has to spend their entire lives, you know,

working to just produce enough food to live. I mean, you can have a bit of poetry in that situation, but again, you see that as a country gets richer, there can be more time available for people to engage in poetry and then share that around and enjoy it as well. So most of the things actually that we do think make life worthwhile depend directly or indirectly on the resources and goods and productivity that we have available to us. And that's why I think GDP is really important as a measure, because it does correlate very highly with many of these other things, both sort of theoretically and empirically if you look at correlation of GDP per capita, which means per person, by the way, which obviously you need to adjust per person to get a measure because otherwise big

countries have just had the largest number to get to get a good measure of the productivity of an economy. But if you take GDP per capita and compare it to say life expectancy or level of education, there's a very high correlation, a very strong relationship between those things, which doesn't prove that one causes the other, although I think that there's a good reason to think that there is a causal relationship there. So overall, I think that there is good reason to think that GDP is quite important, and we'll be discussing it a lot throughout the series of episodes. As such, it's useful to have some idea of how it's measured specifically. So I mentioned that it's based on statistics that are collected by national government

statistical agencies. But specifically, it's calculated by adding up all the expenditures in the economy on final goods and services. So you need to subtract out the inputs that are used. So if someone makes bread and they sell their bread for $2, you don't just add the $2 to GDP, you need to subtract out the inputs that were the cost of the inputs used to make that bread. which say that was $1, so then there'd be $1 added to the GDP for the bread, and then maybe $1 for the flour. But then for the flour,

then you need to subtract out the cost of the wheat, to make that flour. Maybe the cost of the wheat is 50 cents, and the flour is another 50 cents. So the total value added to the economy would be 50 cents for the wheat, 50 cents for flour, and then a dollar on that for the baking of the bread,

making $2 in total. That's just a simplified example to give you the idea. So the idea is you add the what's called the value added at each stage in the production process of goods, services and then add that all up and that's the value of GDP. So that's one way of calculating GDP through expenditures. You can also do it via income. So you just add up everyone's incomes in the whole economy, be that through wages, salaries, profits, interest on money or royalties, whatever

other source of income you have. Add those up over everyone in the economy. It should be the same as all the expenditures because, of course, one person's income is another person's expenditure. Someone gets money, someone has to pay it. So these are two different. approaches to calculating GDP, expenditures and incomes, and they should give essentially the same value. And that's the basic idea of how it's calculated. So basically what we're doing when we're comparing GDP per capita, remember per person, between countries, is just taking the total market value, the total value of everything multiplied by its price, and adding that all up, and then dividing by the number of people in the country, and then so seeing how much resources or goods and

services they have access to, which is a measure of their standard of living, but also more directly, it's a measure of the productivity of the economy, how much value it's producing for each person in it. And so that's why it's used as a measure of development. How developed is the country? Well, how much is it producing in terms of goods and services for each person in it? So nominal GDP is just given in the domestic currency units for a particular year in a particular currency. So in my country, Australia, we might measure GDP in Australian dollars for the particular year. Real GDP, however, is much more useful because it adjusts. for the effects of changes in prices over time or controls for inflation.

So generally you want to convert everything to a common currency, so US dollars is commonly used. As I mentioned before, there's also an adjustment for purchasing power parity, which is the international dollars, or just expressed as US dollars often, because one international dollar equals one US dollar, that's how it's defined. But market exchange rates are not used, because market exchange rates only reflect the value of goods and services that are traded across international borders.

but many things aren't traded across international borders and so their value is not properly reflected by market exchange rates. So instead, what's used, the purchasing power parity rates, which is basically an attempt to take a basket of goods that represents what people buy in different countries and then figure out, well, how much would it cost to buy this in the local currency in India, local currency in Japan, in Australia, in the US, etc., and then use that to work out what's the exchange rate

between one currency and another, but basically how much you would have to, by working out the cost in the domestic currency of the same basket of goods, that gives you an exchange rate. And that obviously needs to be adjusted every year as different currencies change value with respect to each other, and different countries have different inflation rates. So when I'm talking about GDP figures henceforth in this series, I'm always talking about real GDP, so it's always adjusted to, for changes in inflation. So most of the figures that I give are for 2011 US-I-Nus.

international dollars. So again, they're adjusted to US dollars using exchange rates from purchasing power parity. So this attempts to adjust for differences in purchasing power between different countries and effective inflation and, of course, differences in national currency by converting to the common currency. So those are all issues, and for each of them, there are measurement problems, because the measurement of all of these things is imperfect. And I'll talk a bit more about that later on, but this is the best we can do, this is the best data that we have available, And I think we can measure GDP more accurately than we can measure a lot of things. Although that's not necessarily saying that much, but it's saying something, I think.

So that's the basic metric that we're going to use in this series to determine how developed a country is. It's real GDP per capita. And there are certain particular cases when that's not reliable, especially for countries that have very large oil reserves, and we'll again talk more about that later. But generally, that's the measure we're using to determine how developed a country is. Now, before going forward, I want to say a little bit about the difference between rural and urban extreme poverty, because they are qualitatively different, at least in some ways.

In some ways, they're sort of similar. But the point here is that absolute poverty looks different in urban and rural settings. So in many middle-income countries, particularly many countries in Latin America, but also there's various countries in the Middle East and Southeast Asia, as we'll see in a moment, that have, for which this applies. as well, most of the poor live in slums in large cities. So in these countries, the people living in extreme poverty aren't necessarily in traditional villages, living in huts, but rather they're living in slums in the very large cities. In the very poorest countries in the

world, it's the opposite. So these countries don't have very large cities, or they might have one, but still many of the poorest people are living in traditional villages, or some even, in some cases, hunter-gatherer lifestyles, although the numbers they're quite small, mostly living, supporting themselves with subsistence agriculture producing for their own needs in traditional villages. So the urban poor, as I said, often lives in slums in major cities, usually the capital city. The housing is very poor quality, often built illegally, so basically tin shacks or wooden shacks built with scavenged or very cheap building materials, often on land that they don't own or have

proper rights to, very crowded conditions. Some have access to electricity, sometimes illegally, sometimes illegally, others don't, often the access to electricity as well as running water is intermittent, so frequent blackouts or instances where the supply of water doesn't work or the quality is poor, and they don't have very good security of utilities or of the rights to their land, because many of them don't have legal, formal access to it. Most of these people make money working in the informal sector, working as day labourers, running small retail businesses, running errands, construction, various sorts of odd jobs that often have, not always, but often have poor job security.

Living conditions are in some sense worse than in rural areas. There's less space, there's often quite a lot of pollution, more crime and other problems, but wages and job opportunities are nearly always better in the cities compared to rural areas. So the rural of poor generally live in small, generally quite isolated villages with poor road access, generally where their families have lived for many generations. They're often very isolated from the outside world, though they do have some access to it. Obviously that varies from place to place. it's not like these people have no knowledge of the outside world, but it's often quite limited.

They typically live in fact-strived housing, generally that they've built themselves. They often own their own land, often not formally, but through traditional land rights, or they might be also working land that belongs to a larger landowner. That's also common. Often the rural poor have the worst access to health and education services, even less than the rural poor, who at least have some access to the better facilities available in urban areas. Also, unlike the urban poor, jobs opportunities, outside of rural agriculture are extremely limited. In some countries, rural dwellers will

work in agriculture during the harvest times and other periods when it's particularly busy, but then during off-season, they'll go and work as seasonal labor in the cities or in some cases in mines or in other places where there's demand for labor. So there's, in some cases, quite a bit of interaction between the cities and the countries, and other cases less so. Across the world, there's generally, well, for decades really been a steady trend of migration, from rural to urban areas. Basically, it's people move to get better access to jobs, education, higher wages, and better access to services.

Again, this must be understood relatively because being an absolute poverty is bad in the country and in the city, but generally the wages and access to services is better in the cities than it is in the country. Child labour is very common in both urban and rural extreme poverty settings, though often in rural settings it involves children working either in the household, they're helping out on the farm or in family agriculture. That's generally seen by the world to be less of a problem. I don't personally know whether that's true,

whether the sort of physical, social, or whatever, other effects on children are worse. Obviously, in either case, they're not going to school. In many cases, families living in extreme poverty can't afford to send their children to school, partly because of school fees, although that obviously varies by country. Some countries have free education. Some have supposedly free education, but a practice still have to pay fees. Others have free education, but you have to have a school uniform or buy books or other things. And even if there are no upfront fees, there's still the issue of the lost income or labor

from the household of having to send your children to school. And often there's, especially in rural areas, there might be a long way to go to get to the school or might be dangerous. There's other factors as well. In urban settings, child labor more often involves formal employment or children working in the family business. Children begging on the streets is quite common, looking through rubbish stumps to look for items, of Valley or doing other dangerous and dirty jobs. That's still quite common for many families living in extreme poverty. Okay, so the next thing that I'm going to do is take us on a tour of world regions. Now, hopefully this is useful. I think it's perhaps better to have a world map

in front of you when I'm going through this. I'll post up a map on Facebook that I made to help visualize and kind of keep in your mind the different regions. and countries that I'm talking about and the sort of trends. So the purpose of this is to give you an idea of what the landscape of different countries is and where people are living in different conditions and what the sort of current situation is. Different people will divide this up in different ways. I have developed a classification scheme, which is a slight modification of schemes that are used by like the World Bank or the World Trade Organization or these big organizations,

international organizations. I'll talk a bit more about these later, the so-called Bretton Woods Instincts, institutions. So I've tweaked it a little bit, and I think that these are actually meaningful classifications in terms of the level of income that I've split it at, and I'll explain what I mean by that. So I think that this is a useful way of sort of mentally categorizing the countries that's not just kind of picking a random number and saying above this, you're developed and below this, you're underdeveloped, that's obviously not very useful. So of the four categories, the first year the high-income countries, or the developed countries,

So, traditionally, these are also called the First World, although there are a few countries in here now that were never part of the first world. I should say that use of the term first world and third world, it's a bit old-fashioned now. It's kind of deprecated language, but you still hear it. This really shouldn't be used in talking about development economics, because originally these were not economic labels at all, or at least not predominantly, but rather they were more political labels. So this language originates from the Cold War, where the First World was the Democrats, capitalist, American-Allied, Western world, so basically Western Europe, you know, US, Canada, or Australia, New Zealand,

and then a little bit later, Japan was sort of added to that and maybe a couple of other countries here and there. And then the second world was the non-democratic, centrally planned communist countries, so particularly the Soviet Union and their allies. So that includes Eastern Europe, China, and then various other countries around the world. And then every country that wasn't part of either the first or the second world

as part of the third world by default. So the third world was just kind of everyone else. And almost all of the countries that were then part of the third world were also poor, many of them having been recently decolonized by the West, or in the case of Latin America, decolonized, but, you know, a century and a half ago. But, as I said, so that's what the first, second and third world terms mean. Now, no one says second world anymore because the communist world collapsed with the fall of the Soviet Union about 30 years ago now, and so that's a defunct term. I have heard the words the term second world used to refer to countries that are kind of not really poor but not really rich.

This is really not a correct usage. The correct term there is a middle-income country, which I'll talk about in a moment. But anyway, so still in common usage is the term developed country and developing country, and that's probably the single biggest distinction that I will make. So a developed country is basically the same thing as a high-income country and refers to a country that has experienced industrialization, urbanization, has an educated labor force. Most of the labor force has moved out of agriculture

and moved into industry and or services, and they have generally a developed health, an education system, a developed economy, financial system, government, all of that stuff associated with a modern developed country. So the threshold here that I said is about 25 to 30,000 US dollars. This is all 2011 US dollars, remember. The thresholds obviously change as inflation increases the numbers. So fitting into this category of about 25 to 30K and above are all of the European Union countries,

with the exception of Romania and Bulgaria, which are kind of just on this borderline of 25 to 30K, and also Croatia. So these are the three newest members of the EU and also the poorest, and they're kind of just on that threshold of kind of developed, but not quite as developed as the rest of it, sort of noticeably poorer than the other countries. Chile and Uruguay in South America are also roughly at the same threshold. So again, they're sometimes considered developed countries and sometimes not. It depends on exactly who you ask. Again, in terms of overall levels of development, they're pretty high,

you know, pretty high levels of life, expect to see in education and earnings and sophistication of the economy and so forth. But noticeably, just a notch down from, you know, the rest of Western Europe and the US and so forth. But so I categorize them as borderline cases. and Turkey in the Middle East. I think a lot of people don't realize how wealthy Turkey actually is, or maybe it was just me, but I often thought of it as a fairly poor country,

and it's certainly poorer than the US or Germany or whatever, but it's actually fairly wealthy. So I put it as a sort of borderline case of high-income countries. So in addition to these countries that I've just mentioned, and the rest of the European Union, plus Norway and UK and Iceland, sort of discount them implicitly as part of the EU for this purpose here, just to simplify things.

You've got Canada, the United, United States, obviously. Panama is also a high-income country. For a long time, it was sort of run by the US. Japan and South Korea and Taiwan, Hong Kong, Singapore, people know that those are quite wealthy countries. Malaysia as well is also a high-income country. It's not as wealthy as the US, but it is quite developed. Australia and New Zealand, most people know are developed countries, a few islands here and there, which don't bear that much talking about. Israel, which again is a little bit poorer, but still a high-income country. And then there's a a set of Arab oil exporting countries. So basically this is Saudi Arabia, Kuwait, Qatar, United Arab Emirates, Oman, and one other one I'm forgetting. These are very rich countries. In fact, some of the very richest in the world, but it's pretty much solely due to their very large oil export revenues. And so these are actually very unusual countries because they've got a sort of a very obvious two sort of split development. So there are citizens of those countries who,

who either have very well-paying government jobs, mostly government jobs, or get state subsidies and have very high living standards. But they often only comprise maybe half or even much less, even like 10% in some cases of the population. Most of the population are migrant guest labourers, who often have very low wages, work in very difficult conditions, have fairly low human rights standards

that are exercised with respect to them and have no voting rights. And so it's a bit tricky to call them developed countries properly because most of the population doesn't really benefit much from the... I mean, they still own better wages than they could, say, from India, where many of them are from, but not really developed country-level wages and political rights and so forth. So there are sort of an anomaly, and we'll mention them repeatedly throughout our investigation

because it's always important to separate them out from any analysis you want to do, unless you're specifically looking at oil exporters, because they skew the data so much because they're so unusual in terms of their situation. It's also worth noting that apart from the oil exporting countries that I just mentioned, every single other rich country is democratic. Turkey's really been going on democratic backslide the last few years, although they've also been not doing very well economically, so they might slip back to a middle-income country if that continues.

And Malaysia's not the most well-functioning democracy, as I understand, So there are sort of minor exceptions, but basically all of the high-income countries are both rich and democratic, which I think is interesting, and we'll talk more about the importance of democracy later on. So to recap, high-income countries is basically the European Union, North America, but U.S. and Canada, I mean. Australia and New Zealand, Chile, Uruguay, and then the rich East Asian countries, South Korea, Japan, Taiwan, Hong Kong, Malaysia. and then in addition we've got the Arab oil exporters, Israel and Turkey. Now, moving down to the next level, middle-income countries. These are countries that are just below the rich countries

and above the low-income threshold. Now, I mark the low-income threshold as about $8,000 to $10,000 per person per year. That gives, I think, useful borderline cases of certain countries, particularly India, and the Philippines are on the borderline between middle-income and low-income, which sort of seems about right for those countries to me. By this classification, almost all of the countries in Latin America, apart from Chile, Uruguay and Panama, which I mentioned, are middle-income countries,

as well as almost all of the former countries in the Soviet Union, apart from those Baltic states that are now part of the EU, so Belarus, Ukraine, Kazakhstan, some of the other stands, as well as China is a middle-income country. Many of the Arab states that are not the big oil-exporting ones are also middle-income countries, So this includes Tunisia, Algeria, Egypt, Syria, Iraq, as well as Iran. It's not an Arab state, but it's sort of next to that.

Also a middle-income country. South Africa is also a middle-income country, and Botswana are above it. Indonesia and India kind of just below that through. So I regard income as a low-income country, but it's sort of just below that threshold, as well as the Philippines. And Bolivia, which is the poorest country, I think in South America, the poorest country in South America is Bolivia. It's just below the minimum income. status. So the way that I sort of think about it is you've got the high-income countries,

and then most of the rest of the world is a middle-income country, including most of the rest of Latin America, most of the former communist countries, most of sort of the Middle East, North Africa, apart from the big oil exporters, and many of the other countries in Southeast Asia, like Thailand and Indonesia, Sri Lanka. Then that leaves us down to the low-income countries, which is basically most of sub-Saharan Africa, as well as, as the poorest parts of sort of South and Southeast Asia. So this includes, as I said, India and the Philippines, just below the middle-income level.

Probably will bump above it in the next few years. But in addition, Pakistan, Mongolia, Myanmar, Vietnam, Laos, Cambodia. So many of these poorer Southeast Asian countries are low-income countries. A few of the Central Asian countries, a couple of Latin American countries like Nicaragua, for example, and Bolivia that I mentioned. So there's a few in Latin America, a few in Southeast Asian countries, Asia, but mostly the poor countries, the low-income countries in the world are in

sub-Saharan Africa. Sub-Saharan Africa refers to Africa, well, south of the Sahara, but usually, basically, you take the Arab countries at the north of the country, so Morocco, Algeria, Tunisia, Libya, and Egypt, and you take them off, and they don't count as part of sub-Saharan Africa. Sudan is kind of a borderline case. It's often counted in South-Saharan Africa, but it does have a substantial Arab population as well. So, Sub-Saharan Africa, some of the countries there are mostly Muslim, but most of them are not. So the North Africa is the sort of Muslim-influenced part of it, and with a couple of exceptions, most of the countries in sub-Saharan Africa are not majority Muslim.

Most of them are only recently being decolonized, so the last 50 years or so. And most of them, really all of them, with a couple of exceptions like South Africa that I mentioned as well as Namibia and Angola. So they're all right at the bottom of Southern Africa. All of the rest of the countries there are low income. So this is where many of the world's very poorest people live in these sub-Saharan African countries. Now, then there's a category that is below even the low-income countries, which is probably the most subjective. I put it as roughly below $1.5 or $2,000 a year.

So at this point, we're talking very low income. So U.S. in 2011 is about, arounded, say, $50,000 a year per person. That's a similar level to Australia and some of the, richer countries in Europe. And then, you know, other Western countries are a little bit lower than that, maybe down to 40 or 30,000 at the low end. So middle-income countries like China, there may be one-fifth of the level of the US, maybe one-third of some of the poorer countries in Europe. Then when we're talking about low-income countries, we're talking about below 8 to 10K. So many of them are like four or five, six thousand a year, something like that. They're at maybe

10% of the level of, you know, the US and the richer developed countries. The very very, the very very poorest countries in the world, the very low-income countries, have incomes of below, roughly, as I said, $1.5 to $2,000 a year. So at that level, you know, you're talking about less than one-twentieth of the income per person of the US or other, of the richest countries. In fact, the very poorest countries in the world have a GDP per capita in the 2011 international dollars of something like maybe $500. dollars. These countries are very hard to get data from for a bunch of reasons. Many of them are war-torn and have very inept governments, but the point is that there's roughly a factor of

100 difference between the goods and services per person available or the productivity available for the very poorest countries in the world and the richest country in the world. And that's even excluding like the micro-states like Liechtenstein and Bahamas and places like that, they're basically just tax havens. They don't really count. I'm just talking about the places that, you know, it's kind of actually have reliable data from. of which I think Switzerland is the richest and the US is the second richest, if you exclude all these tax havens. But anyway, coming back to the very low income,

so these are a smattering of countries that you've probably heard their names before. So Afghanistan, Syria, Libya, Yemen, Somalia, South Sudan, the Central African Republic, Chad, Niger, Democratic Republic of the Congo, many of them are war-torn countries, or where the government is basically non-functional, like in Somalia, for example. Also, North Korea, they have a functional government. It's just a horrific totalitarian regime

that keeps its basically uses its people as hostages to prevent other countries from invading it or doing things that doesn't want. Haiti is the poorest country in the Americas. It's also in the very poor level. Liberia as well, a particularly sad case if you read up on its history.

So many of these very poorest countries, if we exclude Afghanistan and North Korea, Syria because it's in a recent civil war, but many of these other ones that kind of they sort of form a group through the sort of center of Africa. So in particular, Niger, Chad, Central African Republic, South Sudan, Demerica Republic of the Congo. These are all countries that are landlocked, or very close to landlocked,

that are in the middle of Africa, mostly either desert or rainforest. And many of them have quite a lot of natural resources, although I don't think Niger does. But, yeah, they have very poor state capacity and just very little in the way of any industry or development of any kind. So they're basically as poor as you can get and still not be starving to death

between $500 and $1,000 per person a year. So anyway, that brings us to a conclusion of our tour of world regions. So hopefully you have some idea about, you know, the countries that we're looking at and where they are and where the poverty is. So when we're looking at extreme poverty, we're mostly talking about sub-Saharan Africa

as well as sort of south and southeast Asia around India and Pakistan and Burma and so forth. When we're talking about middle-income countries, we're mostly looking at Latin America, North Africa and the Middle East, and also former communist countries like Russia and China and the Ukraine. And in the developed world, that's the North America, the Europe, Japan, Australia and so forth.

Right. Now, the last section of this first episode will be methodological issue. So I need to talk a bit about some of the methods that we go about, trying to answer the question of why are some countries so much poorer than others? We've got a factor of 100 difference between the very poorest and the very richest countries. Why is this the case?

And why do some countries get to a middle income level and then kind of stay there for a while? Why do some countries able to get very rich? How do we account for this? That's the basic question that I'm attempting to answer in this series of podcast episodes. To attempt to answer this question, we need to use different research methodologies. And so I'm going to talk a little bit about that to finish out the episode. So the first set of techniques that I want to talk about our economic models. So an economic model is a theoretical construct. It's basically consists of usually a set of

equations or variables, some sort of logical relationships between them. The model represents a simplified framework to understand some complicated process. Now economic models have been maligned a fair bit since the 2008 crisis with some justification because there were some very bad models that were used to model very risky investments and that resulted in a lot of people losing a lot of money. However, the fact that there are some bad models doesn't say anything about whether all models are bad or whether there aren't some very good models, so the fact that you can have bad models and they can be misused

doesn't really say much about how useful models are in principle. And in principle, I think that they're very valuable. Theoretical models are used widely in many other disciplines, most notably physics, and also theoretical models in economics usually make quite abstract assumptions about the nature of an economy. So in particular, there's a school, the sort of dominant school of research in economics called neoclassical economics. It refers sort of loosely to a basic paradigm in which you follow to answer an economic question.

So usually it looks something like this. The standard approach looks something like this. You define an agent, which could be like a person, an individual, or it could be a country, or it could be a corporation, or something else like that. So you define an agent, and they have some sort of utility function, which is an abstract description of their preferences. I've talked about some of this in previous economics episodes, so I won't go into too much of the details here.

But basically, you define a function that describes their utility, how well-off they are under different circumstances, so maybe depending on their consumption or how much leisure they have or something like that, or their profits if they're a corporation, right? Then you also define a budget constraint. That determines the limits of the time or resources that they operate under. So it might be they have a certain number of hours a week to work or a certain amount of money to spend on goods or a certain number of workers to utilize in their factory, whatever it is.

There's some constraint that limits what they can do, obviously. Otherwise, there wouldn't be much to say they'd just do exactly what they wanted. But there's always a limit. So there's a bunch of constraint that describes that. And then there's a further set of institutional constraints that are established that basically just to find the situation that they're acting under. So, for example, it might be that if a firm's hiring workers, it has to pay them. market wage that's determined by supply and demand and that they can't control, or it might be the particular tax instruments that can be used by a government, or maybe the number of hours

of work that a worker can decide to work at different wages or something. So there's just a set of institutional constraints that are set up that define the sort of situation and what actions can be taken in that situation. Once we've got those things, utility, budget constraint, and institutional constraints, then the model consists, sort of solving the model, consists of finding the optimum behavior of each agent, which maximizes their utility, so makes them as well off as possible, subject to the constraints they're operating under, and also what every other agent is doing. So what's good for me to do might depend on what other people are doing, and so that defines what's called an equilibrium,

which basically just means something's in equilibrium if, given what everyone else is doing, I don't want to change what I'm doing. So if everyone else is driving on the left side of the road, I also want to drive on the left side of the road. I don't want to just start driving on the right side of the road. because that would be bad. But if everyone else was driving on the right side of the road, then the equilibrium would shift, and I'd also want to drive on the right side of the road. So that's an example of a situation in which my optimal action depends on what other people are doing. So the purpose of solving an economic model is usually to find an equilibrium when everyone is pursuing the action that maximizes their utility,

and then that equilibrium is compared to some sort of optimal outcome that maximizes social utility. So there would be an idea of what would be the best thing for everyone to do, What's the thing that actually happens when everyone just pursues their own interest and see if there's a difference? And usually that's a way that economists try to say, well, is a given institutional setup or a given set of incentives a good idea or a bad idea? Is it going to lead to a good outcome or a bad outcome? So this is pretty much the standard bread and butter paradigm that neoclassical economics follows. And I'll talk more about this in episode four where I talk about growth theories. At least all the modern growth theories pursue this approach.

And indeed, if you don't pursue this approach, it's often hard to get your work published. One thing that I did want to mention is that any model necessarily involves simplifications and approximations and assumptions. And that in itself is not a problem. Some people critique economic models on the basis that they make unrealistic assumptions. I don't think that's a good critique because every model makes unrealistic assumptions. If it didn't, it wouldn't be a model. It would just be the thing. Most physics that you study involves highly simplified models until you get to quite advanced levels.

And then even they're often, they're still quite simplified. But it's about what simplifications you make and how you test them. and whether they're appropriate for the situation. So that's the relevant question in evaluating a model, not does it make simplifications that are quite strong and obviously aren't literally true, but are they appropriate to model the relevant behavior and is the model useful to make predictions and give insight

into the workings of the system? Proper economic modeling of economic growth is extremely difficult because there are so many interacting components that we'll talk about, particularly in episode three, on structural change. But this includes things like the financial system, the price system, governance, agriculture, urbanization, capital accumulation, education, healthcare, monetary policy, and a host of other issues,

and they're all thought to be relevant. And it's really unclear how we can incorporate them all and what simplifications are appropriate and what are not. And so there's no one canonical or standard model of growth. Well, this sort of is, but it's a very highly simplified model. I'll talk about that in episode four. So the point is economic modeling is difficult. It's widely used to understand growth, but it's not capable of answering all our questions, because basically the best model would be one that was far too complicated to actually understand. And so we have to make simplifications, and no one can agree on what the right simplifications are, because it's very hard to check them.

So that's the basic problem. Now, traditionally, economic models were the sort of theoretical side of the discipline, and the sort of empirical side was regression analysis. Now, a regression is a set of statistical processes. Basically, you try to establish the relationships between two different variables, a dependent variable, which is like your Y variable, or your outcome variable, and your independent variable, which is like your X variable or your predictors. Now, I talked about this in a previous episode on statistics.

That's episode 79, basic concepts in statistics, if you want to look at that. But it's basically an exercise in line fitting. If I have a bunch of observations between, say, life expectancy on the X axis and income per capita on the Y axis, I want to see if there's a relationship between them, and so I try to fit a line through those points. And that's what a regression analysis is. You can do it in more than one variable, in which case you have more dimensions, and you can do many fancy things with it, but basically that's what it is.

It's a line-filling exercise. The idea being to try to discover the relationships between variables and try to work out, therefore, the causal structure of what affects what. Cross-country regression analysis is basically where you take the growth rates of different countries as your Y variable, so that's the thing you're trying to explain, and then you take whatever else is your independent variables, and then try to construct a model to try to e-referral, explain the differences in growth. The problem with this is that there's been a kind of everything

including the kitchen sink approach to these, to these regressions. So I'm not talking about you put in one or two or three or four X variables, but I'm talking about dozens of them. In one review, they found 145 different variables, X variables, predictors, that were significant in growth regressions. That is that they found a statistically significant effect, so it's more than zero. The problem with this is that a cross-section regression has maybe 100 observations. I mean, it might have a bit more than that, maybe 150, because you might have one for each country that we have good growth data for. But the problem is that if you've got 150 variables, say, and 150 observations,

then you've got the same number of observations as variables, which basically means that you can't estimate anything reliably. You need more data than the things you're trying to estimate. These regressions are generally seen as kind of a failure, or at least they didn't really answer the questions, that people hoped that they would, using all this data and trying to regress on the results. People just got all sorts of different results. One of the reasons for this is because the quality of the data is poor.

So I mentioned that there are problems with national account data, particularly in sub-Saharan African countries, that the data that they collect just isn't very good. One of the reasons for this is because a lot of activity is non-market activity, so people are producing their own crops and consuming them. That's called subsistence agriculture. But because they don't trade it, there's no price. And so you have to just estimate a price or what's called impute a price. And that's just largely guesswork.

And there's also how do you value government expenditure, how do you value black market activity that's hard to incorporate? And there's all sorts of problems with these sorts of estimates. And so they're least reliable, I would say, for these very poorest of countries. So that's one problem. The data is just not very good. But then there's the issue of the specification of your model. In particular, what control variables do you put in and which do you exclude?

When there's 150 choices, then you've got an almost limitless number of points. possible combinations. And to add insult to injury, then you can make transformations of your variables as well, by like taking the square, or taking logarithms, or you can have interaction terms, which is like when two variables are multiplied together, and there's an interaction between them. It just gets ridiculously complicated. There's a further problem that for many variables, there'll be a causation in multiple directions. So, for example, to take education as a classic example of this, it's probably the case that higher education

helps with economic growth. We'll talk more about that in episode five. But there's also definitely the case that the richer you get, the more money you have to spend on education. So causation goes in both directions there. Plus, there's probably other things that affect education as well, that you haven't included in your regression, which is called an excluded variable. And all of these things resulting what's called endogeneity, which is just basically you've estimated the wrong equation. You haven't put the right variables in, because you won't get the right estimates if you don't estimate the right equation.

But no one knows what the right equation is, and there's effectively, infinitely many possibilities. So what some researchers have tried to do is they just try all the possibilities. I mean, that might sound silly, but that's kind of, obviously, they don't literally try every possibility, but they try millions and millions of different specifications of these regressions, putting in different variables in different interaction terms and so on, and they count the number of times each variable is significant, and then see which of them are significant in the most cases.

Now, personally, I think this is a very bad approach, because my view is that what equation we should estimate should be given by theoretical considerations. We should have some fear as to why we're regressing this particular set of variables on the dependent variable. But nevertheless, these attempts to estimate across many different regressions. It's called Bayesian averaging over regressions

if you're interested in looking this up more. My big criticism of this, apart from the fact that it's very a-theoretical, like this result was significant in 98 regressions of 100. I mean, what does that even mean exactly? But the big problem from my point of view is that the results that we get, so one table that I'm looking at now, for example, from a relatively recent paper, the main variables that are significant are just not very informative. So the most significant one was per capita GDP in 1960. This was looking at growth since 1960 up to like

2000 or 2010 or something like that. So what this is saying is that how rich you are at the start is a very strong factor of how much you grow subsequently. But this is already known, and is a basic result from theory in the Solo Swan model that I'll talk about in episode four. So this is kind of a reassuring result, but this is already very well known. That doesn't tell us anything new. The next two significant variables were level of primary schooling and level of fertility. And again, neither of those are really a big surprise, because we expect that basic education and birth rates have an effect on GDP growth rates.

So neither of those are really very informative. We already knew those. And almost all of the other variables that were reliably successful, significant in lots of the specifications are what are called basically regional dummies. So basically just it's saying, is this country in Africa, yes or no? And if it's yes, then it gets a different interceptor, basically a different estimate for its value than a different country. So just reading here, it's Africa dummy, fraction Confucius, fraction Muslim, Latin America

dummy, East Asian dummy, fraction Buddhist. And those just tell us basically that if the country is in this particular part of the world, then it does better or worse than if it's in this other part. of the world. But that doesn't really tell us anything about why that's the case, or what's the factors leading to that growth difference. So I don't personally find these sorts of estimates over different regressions very informative. We'll talk a bit more about this later in sort of future episodes, but I'm going to be taking more of a theoretically based approach, as we'll discuss more in episode four when we talk about growth theories, and then some of the institutional factors.

So unfortunately, using regression analysis, particularly cross-country regression analysis when you look at different countries, doesn't really answer a lot of the questions that we might maybe hope it would. So, a more recent technique is to use randomized controlled trials. This is basically borrowing a technique from medicine and other science more generally, in which you try to reduce bias by testing the effectiveness of some sort of treatment or intervention by randomly allocating subjects to one or more groups. So say you have a treatment group that you roll out the treatment with, maybe that's some

sort of health or education intervention. You do it in one part of the country, and then you don't do it in the other part of the country. You just have some sort of enrollment groups that you don't do anything with, or you just give a survey, but you don't actually roll out the health or education intervention, and then you conduct the intervention, and then you measure the differences how the groups change over time. Ideally, if you select your groups randomly, then there shouldn't have been any initial differences between the two groups, because they were just picked randomly. So for a large enough random size, there shouldn't be any differences. And that removes a lot of the problems of model specification and data quality and endogeneity

that arise as a result of regression analysis, because you don't have to measure all these variables and stick them in your equation or figure out what are the right variables to include. You don't need to do any of that. All you need to do is pick your random sample, have your control group and your treatment group, and then just look at the difference between them. And all of these other problems basically go away. At least that's the theory. There have been some interesting analyses of comparing randomized and non-randomized studies and seeing how much of a difference is there between them. The results are kind of mixed here. It seems that some types of regression analyses are able to produce similar results.

So this is basically when they ask the same question in two different ways. One does a randomized controlled trial, and the other one just does a regression analysis or something similar to that. And they're asking, do they reach similar results? Ideally, you'd hope that they would, because then we could say, oh, well, regression analysis gets the right answers, so we'll use that. Unfortunately, a lot of the time they don't get the same results. Sometimes they do. Obviously, it's more complicated than that, but I don't want to get into all of the details here. it does seem to depend on the type of regression analysis you do. Randomized control trials can be very useful. They've become more popular. In fact, I think it was in 2019, so last year from the year of recording,

the Nobel Prize was actually won by a group of researchers who've been conducting randomized control trials in developing countries to work out effective health and education interventions. But there are significant limitations of randomized control trials. So they're very expensive and time-consuming. That's one issue. But it's not really the biggest problem. I think the bigger issues is that they don't incorporate, what's called a general equilibrium effect.

Now, we talked about equilibrium a bit before when we talked about theory. Remember, I said the equilibrium is kind of about what you want to do given what everyone else is doing. And that's important because in a randomized control trial, you can't actually see that because you're just treating a small number of people. So the question is kind of, imagine if everyone got that treatment, say it's a new education or health initiative or whatever it is, would the effect be the same?

Perhaps the effect would be better, but perhaps it would be worse. A silly example that I like to give of this, but I think it illustrates the point, is if I went to, well, really any country, and just gave a random set of, say, 100 or 1,000 people college degrees, like, don't give them the education, just give them the degree. I don't mean a fake degree. I mean, like, a legit degree from a legit institution, but without them having to actually do anything.

If you did that, that would almost certainly improve their labor market earnings, even though they don't have a lot of skills, but there's still a significant credentialing effect from having a degree. And so almost certainly they would increase their earnings, maybe not as good as someone who actually went to college, but still, I think it would be quite helpful for a lot of people to get jobs that they otherwise wouldn't be able to. But obviously, if you gave everyone a college degree without actually having to get them to attend school, then that's not going to have any effect on overall earnings, because it's just a piece of paper. It doesn't actually, it's only useful if some people don't have it, right? So this is an illustration. Obviously, you wouldn't actually

do that, but it's a clear illustration is, If you actually did that experiment, you would find probably quite a big effect, and you'd say, wow, this intervention was super effective. But then if you were to try to roll that out across the whole country or whole region, you'd find, in generally, Gorillibrium, the effect would be, well, in that case, zero. In other cases, you might have some effect, but it might be lower. There's all sorts of reasons this can occur as a result of effects on prices or incentives or institutions and other things.

It depends on the details, but there are real concerns there that randomised control trials just really can't pick up because they only treat a small number of people in the control group. Another problem is that randomized control trials don't really tell us much about why something works or doesn't work. You just see the result. You say it's basically yes or no. And I mean, you get the effect size. The size is a difference, but you don't get any information about why that was the case, which often is what we really want to know. And also, a last problem is that randomized control trials don't help us to figure out the different factors that affect the outcome. We just get the average effect. So, maybe.

Maybe the treatment only works for some types of people. Maybe the health intervention only works for people of a certain age or with some interacting health condition or whatever. But unless you're specifically testing for that, then you won't see it. And in particular, it's often hard to tell what interventions will be successful in different contexts, because a single trial can only test a single thing in a single context. Just because something works in rural Bangladesh doesn't mean that it will work in the sloth. of Mexico City, because they're very different contexts, and you don't necessarily know

just from the trial about how it will generalize like that. So there are a lot of limitations to these randomized control trials, apart from just sort of more obvious fact that a lot of things that were interested in, like economic policy, are very hard to do a trial like that for. So they're generally better for smaller scale health and education interventions, but can be used for other things as well, such as cash grants and other things. So that concludes us for the first part of our eight series long. episodes. Hopefully you enjoyed that. If you did, I would appreciate if you could give the podcast a like and or review on iTunes or an aggregator of your choice. If you want to support the show

even further, you can also visit our Patreon page, which is patreon.com slash James Fodle, and you can place to support the show, which is from $1 through to $5 per episode, which I greatly appreciate. We've already got a few patrons there, to whom I am immensely grateful. Supporting the show in this way allows me to spend more time making episodes because it allows me to devote less time to other forms of work and therefore more time to the show, and so I'd love to spend more time on the show. If you'd like to see more episodes therefore, you can make a donation via Patreon. You can also make a direct donation on PayPal if you prefer. You can go to PayPal.combe.combe. That's one word, Science of Everything,

and send a donation directly to me there if you prefer. If you would like to leave any comments, suggestions, questions or other feedback, please send me an email at Fods12 at gmail.com. That's FODS12 at gmail.com. In particular, I'd also value your feedback on whether any of the older episodes have not been working for you to download or listen to. I've heard a few people, a few reports that there's been some problems here, but I haven't been able to replicate those. So anyone who has any information on this, very much appreciate you letting me know. Thanks again for listening, and I'll talk to you next time.