The Science of Everything Podcast - Episode 106: Theories of Economic Growth and Development

Episode Date: May 31, 2020

In this fourth part of our series on economic growth and development, I outline the major theories of economic development developed over the past sixty years. I trace the development of such theories beginning with the Harrod-Domar theory, and proceeding through the Solow-Swan model, the Ramsey model, Romer's spillovers model, and endogenous growth theory, in each case discussing their key features, and analysing their strengths and weaknesses. The episode concludes with a brief survey of a range of more recent growth models focusing on modelling coordination failures and poverty traps, emphasising the work of Daron Acemoglu on economic institutions. The Recommended pre-listening is Episode 105: Economic Growth and Development Part III. 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 106, Economic Growth and Development Part 4, Growth theories. I'm your host, James Podor. So, this episode is a continuation, obviously, in our series on economic growth and development. And today we're going to look at growth theories. So these are theories developed within the economics literature that attempt to provide a generally mathematical,

structural explanation for the process of economic growth how it occurs. So I've characterized this episode as focusing on the how, because in the previous episode, we looked at the structural change and sort of the what, the descriptive analysis of the sorts of changes that happen as a result of and through the process of economic development, including changes in agriculture, urbanization, capital accumulation, and rostos stages of growth. So in this episode, we're going to kind of continue from that and put a bit more structure on it by looking at how those processes occur, not just sort of a description of what happened, but trying to understand the causal mechanisms at play that give rise to those structural changes.

And we'll do that through the guise of formal growth models. And needless to say, recommended pre-listening for this episode is the previous episode, Economic Growth and Development Part 3. Strongly recommend you listen to that, as this episode will make a lot more sense. Formal growth models really only started to be developed following, or just sort of before and following the end of the Second World War. And so I'm going to start with a Harold Domar model, which dates to the 1940s, and what we're going to do is step through successive generations of economic growth models, including the Solo Swan model in the 50s, the Ramsey model of the 60s, Romer's spillovers model, which are some of the early indigenous growth models in the 1980s, and then later fully fledged indigenous growth theory models of the 1980s, and then conclude with the discussion of coordination failures in poverty traps, and some of the more recent models have been coming out since the 2000. So the idea of these is essentially that each new generation of models is an improvement upon and in addition to previous models and that they all help us to better understand the process of economic growth.

These models don't really explain income differences between countries. That's something we'll look at in the subsequent few episodes when we look at the why. This is more the how. How does growth happen when it does happen? Not so much as to why does it sometimes happen and why does it not sometimes? happen. So just understand what these models say and what they don't say. But nevertheless, they do give us very important insights that will be relevant for understanding the process of growth and change. So let us begin with the Harrod-Demar model. So this is kind of really the first macroeconomic model

of growth, and it explains growth by positing that an economy has a simple production function, which just means an equation that describes how much output is produced as a function of certain inputs. And in the Harold DeMarke's taken as simply a linear function of capital. So think of that as like machinery and technology and factories and so forth. Linear just means that output is proportional to the amount of capital and that there is a constant marginal product, meaning each additional unit of capital produces the same additional increment in production. That's a very simple model. And it produces, when you sort of run through the math, some interesting predictions.

particular that as you have more investment, that will lead to accumulation of capital, which will therefore lead to economic growth. So basically, in this model, economic growth is all about capital accumulation, investing in new factories, new machinery, new industries, new roads, all that sort of stuff. Also note in this model, there's no particular reason that growth will be balanced, meaning that different sectors will sort of stay the same size. There's also no reason that this model generates full employment. So under this model, which is a Keynesian model, if you know what that means, otherwise don't worry about it. But this model basically preserves a significant role for government intervention,

both to increase the amount of capital accumulation, so that growth will be faster, and also to ensure that there's full employment as a result of the growth process. In this model, the main factor that determines the rate of growth is the savings rate, which is the proportion of income that is saved and then reinvested, in this case, too, just accumulate more capital. So under this model, economic growth just depends on policy, is to increase investment by increasing savings. And this led to what is subsequently being called the big push strategies for growth,

which is basically the idea that what nations need to grow is they just need more investment. And if that doesn't come from the private sector, then the government can do that, or even an overseas government can do that through aid. And so there was an idea of just giving large grants or loans to countries that would allow them to kind of kickstart their growth by increasing capital accumulation, which would then produce more output, which you could then save and then reinvest, and then continually increase, kind of like money in your bank account.

This seemed to bear out a little bit after World War II with the success of the Marshall Plan in Europe that was basically a provision of aid by America to help rebuild the European economies, and there were sort of similar investments in East Asia as well that worked well in, say, Japan and South Korea and Taiwan. However, similar attempts to kickstart growth in many other countries, including Latin America

and sub-Saharan Africa, in particular, as well as India and elsewhere, were much less successful. And this model, the Harrodomar model, is now regarded as inadequate in describing growth, partly because empirically it just doesn't hold up that capital accumulation is enough for growth, also because of its unrealistic assumption of constant returns to capital, which just means that also the Harrodomar model does not incorporate any role for labour, which we know is a very important input. And also, the Haradamar model just treats savings as exogenous, which means it's just set outside the model by some other factor. That's not really regarded as very satisfactory. So all of these

considerations meant that the Harrodomar model was increasingly seen as unsatisfactory and an alternative was developed. And this alternative is now called the Solo Swan model. Sometimes you'll also see it referred to as the Solo model, but it was sort of developed, I think independently by two scholars in the 1950s. Robert Solo, who won the Nobel Prize. And I forget the first name of Swan, but he was an Australian economist who also developed a similar theory. So the Solos 1 model is effectively an extension of the Harrod-Demar model, but instead of having a simple linear production function, linear in capital,

it uses what is called a Cobb-Douglas production function. So remember a production function is just an equation that specifies how much output or production and economy produces, given certain inputs. In the Harrod-D-Mar case, the input was just capital. But in the Cobb-Douglas case, used in the Solis 1, there are two inputs, capital and labour, K and L. Also, an additional assumption is made here that each of the factors individually exhibits diminishing marginal product or

diminishing returns, which just means that as you get more of one of them, holding the other the same, each additional unit has lower and lower benefit. So the idea would be here something like that if you have a kitchen that has, I don't know, five stoves, then if you hold a hire five cooks, well then you're going to maximize efficiency. If you hire a sixth cook, there's probably useful things that they could do to increase production, but they're not going to be as effective because they don't have a stove. And then if you hire a seventh or an eighth cook, again, they're going to be less effective than the previous ones because they don't have the same capital to use. And it works in reverse as well. If you incorporate more

stoves into the kitchen, but don't have the, if you don't have the chefs to use them, well, again, it's going to add some benefit because, you know, more flexibility for the existing chefs, maybe they can cook multiple things at once. But it's not going to be as efficient as if you increase them in a one-to-one ratio. So that's obviously a simplified example, but that's the basic idea that capital and labor complement each other, and so output only increases sort of maximally if you increase them both in proportion. If you increase one without increasing the other by the same amount, then you'll get a diminishing output. And eventually, if you just keep increasing one arbitrarily largely without changing the other

one then with a large enough increase, the marginal additional output will go down to zero. So this assumption is very important because it leads to different predictions, the different results of the model. This model is also a bit more mathematically complicated, but it's not too difficult, so it's often taught in sort of first year macroeconomics courses. So just to explain a little bit about what the theory predicts, the theory predicts that economies will grow if an economy begins poor, it will experience growth as long as it is able to save enough, because once again, just like in the Harrodomar case, the capital accumulation depends on savings, whereas labor accumulation just depends on the population growth rate,

which is taken to be exogenous. The savings rate here is also taken to be exogenous, which is a limitation we'll come back to. But either way, let's focus on the capital accumulation. So the savings rate will, as long as is high enough, lead to capital accumulation. However, the key difference here is that with a fixed labour supply, capital accumulation leads to a diminishing marginal product of capital because we're keeping the labour constant and increasing the amount of capital.

That means that each additional unit of capital will yield less output than the previous one, leading to diminishing returns. So what this means effectively is that you can't just get the exponential increase that you could in the Harrod-Demar model, because each additional additional increase. unit of capital produces less output than before. And so although you can then, that additional unit of capital produces some output, which you can then save part of that output and reinvest it, as the cycle turns around, you get less and less out of it every time. And so eventually, you reach a point where the additional output that you get from an extra little bit of capital

that you've invested, you only get just enough returns on that in order to offset depreciation, because a part of this model that I haven't mentioned, but capital depreciation. That basically wears out over time, either through physical wear and tear or through technological progress that renders machinery or something else obsolete. But whatever it is, capital depreciates with some rate, and that has to be incorporated in the model. So you have to make sure that there's enough return on your capital to offset depreciation. Otherwise, capital's going to be shrinking, going to be going in reverse. The rate of depreciation increases linearly in the amount of capital, because it just depends on how much you have. the more you have, the faster the overall amount of depreciation.

Whereas the increase in output that you get as a result of the extra capital is less than later because remember it's decreasing marginal product. So basically what you get is this line that slopes upwards, but it's a curve, but the slope gradually flattens out. And at some point, if you just then draw a straight diagonal line, that's going to intersect with it. Like if you have a curve that has an initially steep slope, but then flattens out, draw a straight line, The slope doesn't change for a straight line, so eventually it will intersect that curve.

That's what happens in the Solo Swan model. The depreciation curve intersects with the investment curve. And what that means is that's the equilibrium point for the economy, because now all of the savings, all of the investment in that economy is going to just offset in depreciation. There's no extra that's left over to further increase the capital stock, and therefore the capital stock will stay stuck at that level, and further growth will stop. The economy will just stay at that point.

And so this is called the steady state of the model. This is a prediction of the Solar Swan model that if you're just relying on capital accumulation, that will only get you so far until you reach a point where capital accumulation and depreciation just offset each other and then the economy stops growing. Now, the situation is a little bit more complicated if you incorporate population growth. But what we're really interested in is output per worker, not overall output, because output per worker gives a sense of the productivity and the living standards of a country. So if you divide everything by the number of workers, then you still get the same results, even with population growth.

The mathematics get slightly more complicated, but the basic intuition holds that just having more workers doesn't really change the situation. What matters is how much capital you have for each worker. So what the Solo Swan model predicts is that an economy that's very poor will start growing rapidly, but as they get closer and closer to the industrialized wealthy countries, their growth will slow down and eventually stop. The only way you can have sustained growth in the SolarSwan model is through a mechanism that's called total factor productivity. And basically this is just a multiplicative factor that you multiply everything else by in the model. It's taken as exogenous. That is, it's not explained within the model.

You don't say that it comes from this or that. It's often interpreted as representing technology, but that's kind of a loose interpretation because the model doesn't really say that. basically what it means is that if you assume that this total factor productivity, this multiplicative factor increases over time, then you can get sustained growth. So this is often taken as representing technological progress. But also, this is usually assumed to happen kind of the same rate at all different countries. So basically, the way it works is that suppose that we think that technological progress is leading to 2% economic growth every year in the richest countries. So they're growing at 2%. poorer countries can grow much faster than 2%,

because they're growing at the 2% productivity growth rate plus the catch-up growth that they're doing, which means they're accumulating capital and getting closer and closer to the developed world. Progressively, their growth slows down as they accumulate more capital. You get those diminishing returns in,

and eventually the catch-up growth stops, and that's when they converge. The word is convergence. They're converging to the developed world. And at that point, they will only grow at the 2%. the rate of technological progress that kind of everyone gets for free, so speak. And that's the prediction of the Solo Swan model that you would see this convergence,

that poor countries would grow faster, but as they accumulate a capital and get diminishing returns, then the growth would slow down, middle countries would grow less quickly, and the developed countries would grow slowest of all. In practice, this is not what we see. In fact, there's no particular relationship between rates of economic growth and level of economic output. However, that doesn't refute the SolarS1 model because the SolarS1 model only predicts convergence

if different countries have the same level of technology, that's the factor that I mentioned that grows over time, and also if they have the same production function, which essentially means that they are able to transform capital and labor into output with the same efficiency and in the same way. And we know that that doesn't happen even approximately because different countries have institutions that work much better than others

and all sorts of complicated factors that we'll get into in later episodes. So the way that it's usually phrased is that the Soliswan model actually predicts conditional conversions, which means if two countries are basically the same, then the one that is poorer will be able to grow faster than the one that is richer because it's able to experience that catch-up growth. It doesn't say that poor countries will overtake rich countries. in fact, the Solis 1 model never allows for that. All it says is that they'll gradually catch up,

and then the rate of growth will slow as they catch up, and eventually they'll converge and reach the same steady state, at which you only grow at the rate of technological progress, say roughly 2% a year. And there's quite a lot of empirical evidence to suggest that this house held up. So it's difficult to do this, but if you control for various differences between countries and try to just focus on the level of output,

then you do find that for countries with similar institutions, poorer countries grow faster. So one of the strongest evidences for this are the countries in the European Union, which have particularly the Eastern European countries, are poorer because coming out of communism a few decades ago, they have over the previous, I think, three-ish decades, grown faster than the countries in the West of Europe. And that's exactly what you would expect from the Solo Swan model.

And there are other case studies that have been done from this as well. And broadly, they support the model. So this model is widely, it's still, I think, the most widely used growth model and widely taught. And I think it gives us a lot of insights about the idea of catch-up growth and convergence. And this is an absolutely critical lesson to remember because even, to be honest, professional economists I sometimes see making arguments, as we'll see later, that don't fully incorporate the effects of convergence and catch-up growth. that is when countries get richer we expect growth to slow down and so if you're trying to look at the effects of changing policy or changes in international circumstances or whatever else you must make sure that you factor out any changes in overall wealth of a country over time because you expect

countries to slow down we have been seeing this in the last decade or so in china which about 10 years ago was growing at 10% plus a year but nowadays has been growing up much slower, more like 4% in real terms per year. And I expect it to continue slowing down as it gradually catches up to the west. This is also what happened in South Korea and in Japan and Taiwan and many other places that initially grew very quickly. Then their growth slowed down as they gradually caught up to the west and then they reached maturity as a developed country and had more modest levels of growth. Despite its successes, the Soliswan model still has important limitations. And one of them I mentioned before, the savings rate is taken as exogenous, just like it is in the Herodema model.

And that's not taken as being very satisfactory because savings are not just determined by, they're not like the weather, they're determined by something outside. They're determined within the economy as a result of consumers and corporate and political decisions. And so really a good model should incorporate savings into the model. In order to do this, you need to have significantly more sophisticated techniques. and this is what the Ramsey model from the 1960s does, which we'll now move to talking about. This is also called the Ramsey-Cass-Coopman's model, and it differs from the SolarSwan model

in that the choice of consumption is now modeled as maximizing consumer utility at a point in time. This ensures that the savings rate is endogenized, and that just means that it is incorporated into the model and not just taken as a parameter that you just stick in and just comes from wherever. This model is much more mathematically complicated than the Solo Swan model and is therefore not usually taught at undergraduate level.

But basically it involves specifying a household utility function for every household. In fact, the idea is you take a representative household that represents the entire country and then just sort of multiply that out by how many households are in the country. That requires its own assumptions, but we're not going to get into that here. So you take a representative household, you specify a utility function which says, well, for a given amount of consumption, this is how much utility the household gets. and then you also have a budget constraint for the households, which basically represents their wage income

and any returns from investments that they have, so that's their savings, and then you maximize that utility subject to that constraint. So basically, what choices will the household make in terms of balancing consumption now for consumption in the future, subject to the income that they have, which is income from their labor and income from the investments and savings that they have?

Obviously, if you save more now, that gives you more income in the future, but at the cost of less consumption now. So there's a trade-off there. And how that trade-off is balanced, of course, depends on the particular preferences of the household. So in these models, you have a parameter that represents basically how patient the household is. So that's the discount rate. How much we're willing to sacrifice now for a benefit in the future. If you have a high discount rate, that means that I require a lot of benefits. benefit in the future to give up anything now. So an example would be how much money would I be

willing to pay now to get $100 a year from now? Obviously, I'm not going to pay more than $100 because that would be pretty dumb. I could just keep the $100 and get $100 in a year. But presumably I'm going to pay more than $1, for example, $1 now to get $100 next year. Well, yeah, I'd pay that. I'd certainly pay more than that. I'd pay $10. I don't know exactly how much I would pay these days, probably nearly 100 because interest rates are so low. The point is, though, the more patient you are, the more you would be willing to pay for $100 a year from now, the less patient you are, then you're willing to pay less and less because you're requiring more benefit to yourself in the future in order to give up that consumption in the present. So that's a key parameter

that goes into these models that determines the rate of savings. Other parameters that go in include the depreciation rate, which again is taken to be exogenous. and the population growth rate also take it to be exogenous because that's kind of its own thing. And you get a prediction about the level of the savings rate. We won't go into the details of that. It's not actually that important. I think one of the lessons of the Ramsey model is that although you have a lot more mathematical complexity, the actual insight that you get into the growth process is not that much greater. One result that is important is that in the Ramsey model you can prove that the savings rate is Pareto optimal.

What that means essentially is that the savings rate that is produced naturally by households making decisions to maximize their own utility is the best savings rate that you could get in the economy. Saving more would actually make you worse off in the long run basically by giving up too much now to get not enough in the future. And then saving less than that would be not giving up enough to get enough in the future. So again, there's assumptions that go into that that we're not going to get into here. The point is, though, remember I said in the Howard Demar model that there's no result from that that, that, says that the growth would be optimal or that the savings rate would be optimal or that you'll even have full employment. Well, actually, the Ramsey model does prove those results. It shows that growth will be optimal and, well, a corollary of this is that you would have full

employment. And basically, it's doing that by incorporating a lot more what's called micro foundations for the growth process. And that means that instead of just assuming very abstract production functions that represent like a whole economy, you actually model the process of consumer choice, like individual households making decisions about how much to save and how much to invest. And that's a much more fine-grained level of analysis. And when you get that, you're able to prove that the growth, the level of savings is optimal. And that's an important result because it follows from that that taxation of returns to investment, for example, will lower the equilibrium level of capital and therefore make the nation poorer.

One question that naturally arises from that is are different levels of taxation, both formal taxation and other distortions that have the effect, have similar effects to taxation, are different levels of taxation able to explain differences in income levels between countries? And it seems like it may be able to explain some of it, although you'd have to assume very large levels of taxation to be able to explain the income differences,

because remember, income varies about a factor of 100. Income per person varies by about 100 times between the richest and the poorest countries today. So you have to have pretty big tax levels to explain. explain that. So it's not thought that different levels of taxation. You remember this includes kind of formal and informal taxation, which would include like corruption and other things. You'd have to have pretty high levels to explain that. But nevertheless, it is thought to be a factor that explains different income levels and different growth rates in different countries.

Fundamentally, though, the Ramsey model doesn't really add that much that we don't get from the Solis 1 model, apart from being able to now say where savings come from. This general approach, however, is basically the standard in, well, really all of MacRae economics these days is to take a micro approach. That is instead of just saying, well, the economy behaves like this, we actually start with consumers and firms and say, well, here's their utility, which represents, you know, their different preferences. And then here's their budget constraint, which represents the resources they have available. And then they're going to optimize their utility. So basically do as well as they can, given their budget constraint. And what predictions do we get?

What results do we have when we sort of churn through the math of that and consider the different outcomes? And so that's the broad approach that's taken by most of the models that follow. And so it's very important methodologically to kind of understand how this is also generated. All right, let's move now from the Ramsey model to the, what I'm calling the Roma Spillovers model. Sometimes it's just called the Roma model, but Roma has different models. So this dates to the 1980s, and it's an early example of what is called endogenous growth models. Now, the basic idea here, remember that in the Solo Swan model, and the Ramsey model too, really, What we have is an explanation of catch-up growth.

So capital accumulation, as a result in from savings, leads to an increase in output, but that is diminishing because of diminishing marginal product of capital, and eventually you reach the steady-state point at which you sort of caught up to the rich countries, and then you don't continue to accumulate capital. At that point, the only source of long-term growth is due to increases in the total factor productivity, which we loosely interpret as changes in technology, technological improvements. But this is not incorporated or explained in the Solar Swan model at all or in the Ramsey model either. And so it's left as exogenous.

Exogenous, again, meaning it's just outside the model. It just comes from somewhere. We don't explain where it comes from. While the Roma Spillover's model attempts to endogenize that productivity growth, and that's why it's called an endogenous growth model. What that means is it's attempting to incorporate the effect of technological progress into the model. and these were the first models to try to do that. So the Roma spill load those models

is a fairly simple way of trying to do that. It basically just assumes that the total factor productivity factor that I mentioned, it just grows at some rate proportional to the capital stock. One of the ideas behind that that's been very influential is the idea of learning by doing. What that means is essentially that as you do more of something, you get better at doing it. And so the more of something that you do,

the more of something that's done in a country, the cheaper it can be done and the more productive you get. So that makes a lot of sense for a lot of industries that basically, while it's inefficient at the start as people are learning techniques and people are developing skills and they're figuring out how to arrange their factory or schedule their workflow and all the other things that need to be optimized. And as you do that more and you practice and you increase your scale, then you get better at doing it.

So that's the basic idea of learning by doing. The other side to this is that there's an externality to it. Remember an externality is either a, cost or a benefit that affects other people other than those who are engaged in the transaction. In this particular case, what it means is that suppose I set up a factory and start making cars, I get better at making cars the more of them I make, and that's learning by doing. But the idea is that it doesn't just affect me. It affects other car manufacturers as well, and therefore it's an externality. It doesn't just affect your factory or your workers,

but has a broader effect on the industry. Now, exactly why we would think this is a bit controversial, and there's much that's been written about this. The evidence for these sorts of learning by doing externalities, it does exist. We'll talk more about this later, but it's not foolproof. There are issues with it. It's quite hard to measure.

One of the ideas as to how this happens is that workers are trained in one factory or one shop or one firm or whatever it is, and then they move to another one and bring the skills with them. And so that makes sense. Another is that management moves around between firms or between factories as well. And so they take their techniques with them. But also there's just a production of knowledge, people publish in trade journals or discuss at conferences or whatever else, things that they've learned, things that work, things that don't work. New techniques are developed and published.

New technology is developed and spread throughout the industries. People talk to each other and trade ideas. So there's all sorts of formal and informal ways that we think that this kind of learning by doing can happen. Regardless of the exact mechanisms, what you can do is you can incorporate this in the model and then kind of see what happens. And in the case of the Roma spillover's model, when you assume that technology grows at a rate that's proportional to the capital stock, what you actually have is a model in which you get, well, in this particular instance, you get constant returns to capital. So you remember in the Haradamar model that they had constant returns to capital and therefore you could just increase output without limit as you just accumulate more and more capital. But the Solis 1 model said, well, hang on a minute. that's not realistic because there's capital depreciation and then there's diminishing returns to capital.

So you're going to get less and less out of each traditional unit, and so eventually you're going to reach a balance point at which you can't accumulate anymore. Well, Roma kind of flips that on its end again and says, no, actually back to Harrod-Demar, except with a twist, because Roma's model has labor in it, whereas, remember, Harold DeMars did not have labor in it, and that's an important difference. And the other thing is that the constant marginal product is coming from a different source. is not coming from capital directly, it's coming from the effective technological progress that make capital more useful over time. So it's a different conceptualization of what's happening. One of the effects you get from Roma's model is a scale effect, which means that growth accelerates as you get more and more labor.

Remember that there's labor in this model as well. This is common in many endogenous growth models, which we'll discuss in a bit in a moment, and is an interesting prediction because it seems to say that with population growth, as there are more people around, then growth accelerates. It's not clear whether that's realistic or not. For example, the United States has grown at about 2% in real terms per year for over 120 years, maybe slowed a little bit in the last few years. That's a bit controversial.

But at any rate, the overall rate of growth has been pretty consistent, even though the population has increased many times over that century. And likewise for other countries as well. So also, we don't see that large countries. countries grow faster than smaller ones. China doesn't grow faster, you know, than Taiwan, or hasn't, historically at least, just because it's a lot bigger. So there seem to be issues there with this assumption. On the other hand, perhaps that country is the wrong level of analysis to look at here, because if there are externalities, they may not just be constrained within the borders of a single

country. Perhaps what we really should be looking at is kind of the whole world, or at least the world that's integrated with the global economy, which these days is basically the whole world. The point is that if we look over centuries, we actually do see that as the world population has increased, the rate of economic growth has also increased. So the world population is larger now than it's ever been, and the overall rate of economic growth, now, at least compared to past centuries, is higher than it's ever been. So perhaps there is something to this scale effect. So that's still a bit controversial, and people debate about how relevant these models are. But Roma's spillover's model, although it's fairly simple, does have important insights. One is this notion of learning by doing externalities that can lead to sustained growth over time.

Another is this scale effect, which again controversial. And the third thing is that in Roma's model, there's no proof here that growth rates are optimal because firms won't necessarily consider these external benefits, the externalities of learning by doing on other firms when they're deciding how much to invest. So again, we're kind of going back to the days of the Herodemar model in which the amount of investment wasn't necessarily optimal. Ramsey showed if you endogenous savings that it was optimal. Now with Roma's model, he's shown again that because of these externalities, it's not necessarily optimal. So this illustrates the point that, of course, whether something's optimal or efficient or otherwise depends on the assumptions you make.

And we gradually add more assumptions and make the models more complicated and hopefully more realistic over time. So that's the Roma spillover's model. Let's now turn to modern endogenous growth theory from the 1990s. This theory has developed from Roma's early work and has made much more complicated models. I won't discuss them in detail here. There are different types of models that try to describe how technological progress happens over time. And there's basically two main categories of these models. Either they're models that are based on changing the variety of goods that are produced.

These are called expanding variety models. So instead of just one good that's produced, there's many different varieties of goods that are produced. The second variation describes a hierarchy of improved quality of goods. These are called Shumpeterian models. Both types of these models exhibit increasing returns to scale, that we mentioned before with the Roma model, and also have the potential for non-optimal growth paths for the same reason because of these externalities that we discussed in the Roma case. The big problem with these models, and the reason why I'm not a big fan of them, and they have been criticized more recently, is that they involve many specific assumptions about the structure of markets and the incentives

in those markets. When you try to construct a model of how the range of goods expands over time or how you have different quality hierarchies of different types of goods. Clearly, that is a fact that happens in and does occur over the process of economic growth. You get more goods and more different types of goods and higher quality goods. But the specific assumptions you need to make to make that precise are often quite rigorous and specific and hard to test empirically. And different models make different predictions.

And I would say that overall, there's not been a very clear consensus within the indigenous growth literature as to what models are more appropriate or what we can really learn exactly about the process of growth, other than the fact that these sort of external benefits or externalities do matter. That's a clear result. Exactly how much and in what context is less clear. Also, as noted, indigentonious growth models still don't say much about the causes of income differences between countries. And really none of these

growth models do, although the last set that we'll talk about in a moment in coordination failures has more to say about that. But most of these are about how growth happens and not why it happens in some places rather than others. So I won't hold that too much against an endogenous growth theory. So that's all I really have to say about adogenous growth theory. It often, I think it was high on promise and small on delivery, because it's really not clear what we've learned about how economic growth occurs that we didn't already know from, you know, Roma spillover model in the 80s. We've got some complex models, but they're very hard to test. Unlike the solo model, which is actually quite easy to test and has been, I think, very well verified. And the Ramsey model

being an extension of that, which endogenizes savings, kind of incorporating micro-macro theory. That's a very significant development. Indogenous growth theory, I'm less enamored with, but you will, of course, find people with some of different opinions. Let's move on then and talk about the final category of growth theories that I want to discuss today, and that is, I'm loosely calling coordination failures and poverty traps. Now, the basic idea of coordination failures of poverty traps is as old as the Harrod Demand model, but recently there's been a lot of work in the last couple of decades since the early 2000s. I think there's been a significant increase in the amount of focus on these types of models as being important explanations of how growth occurs

and also of trying to explain growth differences between countries. So what do these models have in common. Unlike the previous cases, that there's nothing very specific holding these models together. They're kind of a grab bag of different things, which is one of their limitations. There's no overarching framework. However, there are commonalities. And the basic commonality here is that they attempted to explain growth in income differences and changes over time in those by the operation of something that's not working properly. So pretty much all of the previous models assumed that things were efficient, or at least could be efficient in principle, so that firms were working efficiently and that workers could be hired and fired effectively, and that there was, you know, perfect

information and all the other typical, and all the other traditional assumptions that are made in the simple economic models, which are necessary to sort of get the model started, and then you can relax those assumptions. But I guess these models are coming to the point of, okay, it's time to relax some of those assumptions and see what happens. And of course, the result depends on which assumptions you relax and exactly what you assume instead, because you always have to have some assumption. So again, there's no one result here, but there are some very interesting insights that we can gain into the growth process by looking at some of these different models.

And I've kind of divided them into broadly micro-level models and macro-level models, depending on whether they deal with the incentives-facing individuals and firms, on the one hand, or whether they mainly deal with political and macro effects like taxation, and aggregate demand and things like that. Let me start, however, by talking about, it's not even really a model, it's sort of a general framework, and this is the idea of multiple steady states.

I think one of the most important concepts to really come out of this coordination, failures and poverty traps literature, just like Indogenous Growth Theory, had their learning from doing externalities. This set kind of has its multiple steady states, which is, now the basic idea here is that a steady state is a condition that is,

stable in some sense. So we saw steady states in the SolarS1 model where you had a condition in which the extra saving that you, the extra output that you get as a result of capital is exactly enough to offset capital depreciation. And at that point, the pluses and the minus is

exactly offset and you reach a balance. And therefore, the overall level of capital doesn't change and you're in a steady state. It doesn't mean that the steady state can never change. It just means something outside the model, like a change in one of the parameters or a shock, for example, a war or a new technological development, something outside has to come in and change that. So that's what a steady state is. One of the goals of economic models is to find steady states because that produces a prediction

as to what we would expect to see, at least in the long run. Again, it could take a long time to get to the steady state. But some models have not just one steady state, but multiple steady states. The SolarS1 model only has one steady state. Think of a diagonal curve. like a 45-degree angle going up along your y and your x-axis, and then imagine a curve that starts out with a steep slope, but then gradually tapers off, like a blade of grass bent over, right? It sort of goes up and then and then tapers off, and then eventually that'll intersect your diagonal line, right?

And that's your steady state. That's when the two things are equal. In models with multiple steady states, instead of having a line that looks like a bent blade of grass and that just intersect once, you can have lines that look like snakes, going up and down and squiggling back and forth along the 45-degree line.

And basically, whenever you intersect the 45 degree line, that's going to be a new steady state. It's a little more complicated than that, but just to simplify things. Remember, because the intersection represents the balancing of different forces. It's sort of like it could be the capital accumulation and the capital depreciation. When those two intersect, they're at the same level. They balance. You get a steady state. But we can extrapolate from that.

It doesn't have to be just in that particular context. It can kind of be anything. Whenever we have two forces acting, one increasing, one decreasing something. Whenever they balance, you get a steady state. And many models can have multiple steady-states, which means basically there's two different ways it can go. It can go one way or another way. And in particular, often what we're interested in in terms of economic growth is that you might have a steady-state corresponding to high-income and another state-states-state corresponding to low-income. And that's a very interesting phenomenon, in which case that the model actually has two possible outcomes.

So you can explain both poor countries and rich countries in the same model, without a very interesting phenomenon. even needing different parameters. They just end up in a different state of state. That's a very interesting result. Now, that's all very abstract. Let's try to make this concrete by sort of explaining how this might work in particular cases. And let's start with some of the micro models. Many of these models are taken from Asimoglu's textbook on economic growth. Some of them also come from other authors as well. But again, many people have published similar sorts of models and explored different things. So I'm not going to focus too much on exactly the precise form of the model, just the general idea behind it to give you a sense of how these things work and what

insights they can give us into the process of growth. So let's start with some of the micro models, and one that models what happens when you have increases in agricultural productivity. Basically the idea here, we talked about this in the previous episode with structural change and urbanization. In order to have enough workers to man the factories and to produce output, you need to have enough food to feed them. And to have enough food to feed them, you need to have enough agricultural productivity so that the agricultural workers can provide a surplus that then allows freeze up labors to go from agriculture to work in the factories. That's essential for urbanization and for industrialization. So under this model, you can show that if you have higher

agricultural productivity, that allows a larger proportion of workers in manufacturing. Okay, well, but then if you combine that with learning by doing, remember, that's when you do more of something you get better at it. That's the externality that we talked about in the Roma spillovers model. if you have learning by doing in manufacturing, that results in higher rates of growth. So basically the idea here is that high agricultural productivity allows more workers in manufacturing, which allows more efficient manufacturing, which allows more growth. So there's a potential for multiple equilibrium here. If we have low agricultural productivity, we therefore have few workers in manufacturing,

we therefore have very little learning by doing, small scale, low tech only, and therefore low efficiency and output. On the other hand, if we have higher agricultural productivity, then that allows for more workers in manufacturing, which lasts for more workers by doing, which allows for higher rates of growth. Now, this may or may not be an issue, an instance of multiple steady states, depending on how you construct a model, but one instance in which it would be, for example, is if there was a feedback effect between high industrial productivity and high agricultural productivity. So suppose that in order to get higher agricultural productivity, you needed to have

factories to make, produce fertilizers and pesticides and, you know, to the industry to build roads and the transportation and railways necessary to transport the goods and to make machinery needed for the factories and to engineer new types of seeds. You see that in this case, basically high agricultural productivity supports high industrial productivity, which in turn supports high agricultural productivity. And so if those two go together, that's going to be a stable, steady state. But then the other two opposites also go together. Low agricultural productivity goes along with low industrial productivity, which then in turn supports low agricultural productivity because you don't have all of those inputs that you need.

So this would be an example of a model in which you'd have two steady states. So I think the key takeaway from that model is just the importance of agricultural productivity. And that if you can get a boost in agricultural productivity through, well, various means, that can perhaps significantly accelerate the process of industrialization. And we had talked about that previously as well. This seems to have been what happened in Britain. In the lead-up to the Industrial Revolution, there was first revolution in the efficiency of agriculture. And that appears to have played a role in facilitating the Industrial Revolution.

So I think this model is actually quite, quite useful. Now let's look at another model. This one is a bit different in that it shows how if we have inappropriate technologies, and this is something that's been discussed in the literature for some time now, that that can lead to significantly lower amounts of output in some countries to another's. So this is not exactly a multiple steady states model. However, it is a model that attempts to explain cross-country income differences by saying that the same technology, and a classic example of this would be something like a heavy plow, which, although it's a not, technology, it works well in the rich, dense soils of Northern Europe, but it doesn't work so well

in same climates. You can have that same technology, but in some countries it's just much less useful and hence inappropriate, or at least less appropriate. And therefore, you can have the same technology available in both countries, but just much lower output in others. That might seem like a fairly obvious result, but constructed in a formal model, you can actually show how you can have quite large differences resulting from these effects. And so I think that's a model. And so I think that's a model to bear in mind as well. And it's not just agricultural effects. There are other technologies that may be harder to use, for example, in different climates or in countries that are more rural or in countries that have different institutional environments. So this is an example of what's

known as a poverty trap. The idea of a poverty trap is basically that something about being poor makes it impossible for you to do the things, or at least very hard for you to do the things needed to get out of poverty. So the inappropriate technologies would be one example of that. You can't adopt the new technologies because, or you could adopt the new technologies, but they wouldn't actually be very helpful because they're not very appropriate to your environment. But in order to get the environment to be appropriate, you need to adopt the technologies in order to therefore improve your efficiency. You see how there's a sort of a catch-22 there. There's a bind that prevents you from doing that. We'll see more examples of that as we go forward and look at

more of the models. So so far we've seen the importance of agricultural productivity and the importance of having appropriate technologies. Let's look at another one, which is a model of rule to urban migration. This is a very interesting model because the basic idea of this model is that you can show that if you have a different levels of effectiveness in enforcement of contracts and other institutions in rural areas compared to urban areas, then that can cause urban migration to be slower than it otherwise would be or that would be optimal. This might not seem obvious, but the basic idea here is that you might think that it's easier to monitor transactions and to detect cheetahs and to enforce contracts in a, in a real situation.

rural environment in which people know each other and there's a tight-knit community and there are traditional mechanisms for enforcing these sorts of things compared to an urban environment in which those traditional institutions don't work or aren't applicable because it's a new growing urban environment in which it's much more depersonalized you don't know people and it's much easier to sort of run off or escape if you cheat someone and if that's true then actually the rural sector has an advantage over the modern urban sector not in technology but in enforcement of contracts and related institutions. If this is true, then certain proportion of people are going to be less inclined to move into the urban sector because there are other benefits basically to just staying in the rural

sector. This can give rise to a dual economy, which I think we've discussed before, the idea of dualism, where alongside each other, generally in the cities, you have fairly wealthy, efficient firms using modern techniques, and then, you know, in a shack just next door, you've got traditional workers using very primitive techniques with low skill and low capital intensities and low efficiencies. And they kind of go along with each other. The strange thing is understanding how those low, low skill, informal economy firms or just single workers are able to compete with the modern sector because oftentimes they're producing similar goods. They compete with each other to some degree. That's still an open question. But one of the results from this is that

if you have a lag of rural to urban migration, then you can get this. this sort of dichotomy, this dual economy. The details of that I won't get into because it gets a bit technical, but I think it's interesting to note that this affects of basically different institutional enforcement means that it's more difficult in urban settings to enforce contracts and other things like that, which can then slow the growth of these large, modern industries, and then basically trap a lot of people in low-skilled, low-productivity traditional sectors. Now, another very interesting model, which is a very clear example of a multiple

equilibrium or multiple steady states model is one in which there are searching and matching costs between workers and firms. And this is a very important complexity to add to our economic models because of course we know that all workers aren't equal and firms require slightly different skills and it takes time to match a worker to a job. Even in fairly unskilled industries or even if there's only a small amount of skills, this can be an issue. But it becomes more of a problem as workers potentially get more training and as employers want to adopt more modern techniques and that require trained workers. So there's a kind of chicken and egg problem here

because before a firm can adopt modern techniques, modern machinery, so forth, it requires workers with the skills. This might be as simple as reading, but as you, again, get more complicated, there's more specific training and skills that are required. You're not going to be able to do that effectively if there are just no workers who can do the work that you need.

On the other hand, there's no point in getting the training to do something if there's no one paying, if there's no employers available to provide a return to that. So there's kind of a coordination problem here. This is actually specifically a case of multiple equilibrium rather than multiple steady states. The difference is technical, and I'm not really going to worry too much about it, but just be aware that there is a difference here.

But in this case, you can have one equilibrium in which there are low skills, so workers don't get the training, and therefore because you can't find anyone with the training, then employers don't provide or don't use the modern techniques that would require their training. and therefore the economy will be poor. The second equilibrium is one in which their workers are highly trained in skill, and then you have employees using the modern techniques and employing those skilled workers.

Once you're in one equilibrium, there's no real reason to move to the other one, because things are balanced out with each other. If I'm a low-skilled worker, there's no real reason for me unilaterally to get better skills, at least, you know, at the margin, it's not going to be that effective because there's not many people who employ that. And conversely, if I'm a firm, it's hard for me to use modern techniques, because I won't have the labor that's necessary. Of course, there are ways of getting rammed.

It's like importing labor from other countries and so forth. But nevertheless, the existence of these searching and matching costs can significantly impede the ability of firms to adopt more modern techniques, which, again, is one of the key mysteries of developing countries, why inefficient and old low-skill techniques are able to continue to exist, even alongside sometimes much more modern techniques. So we've seen some reasons for that. One reason that we spoke about in the agricultural episode is because modern techniques

are potentially more risky, and that's particularly relevant in agriculture, because the farmers may be quite risk-averse as a result of living on the edge of subsistence. Another reason we just talked about is because of inappropriate technologies, which make the technologies less applicable. Yet another reason that we just discussed here are searching and matching costs between workers and firms. Let's move on now to the macro or political level models, still within the coordination failures and poverty traps section.

And these are at a sort of a bit more higher, a bit higher level of abstraction at the moment. And so let's talk about a model of aggregate demand externalities. These are really big these days in these sorts of models. And so I'll discuss what it means. Basically, aggregate demand just represents the total amount of demand in a economy for goods and services. so like what everyone wants to buy basically the total amount and that can depend on various factors

but the point is that if aggregate demand depends on income which it does which in turn depends on production which it does because you know someone's production is sold and then becomes someone else's income then when you're deciding whether or not to make an investment whether or not that's going to be an effective or profitable investment depends upon whether other businesses in different industries also undertake investments at the same time or at a similar time. So this gives rise to Maltzbole-Kulibrao. So again, the basic idea here is that if I build a factory in a poor country or a city in a poor country, that's going to produce a certain type of good.

Let's say I'm thinking about making a shoe factory, but it could be an anything factory or other industries as well providing services. Now, that's going to increase people's income because I'll employ some people and pay them wages, plus ultimately I'll need to buy inputs that, again, I'll buy from people, and I'll increase the product, the output of that country, because now I'm making shoes. Let's say that I'm building a shoe factory to make shoes more efficiently and effectively instead of lots of people making a few shoes in the old handmade way, which is much less effective. So I'm increasing people's income and I'm increasing their productivity of the economy.

The problem is people don't want to buy just shoes. They want to buy lots of other things as well. So although I'm increasing people's incomes, they don't just want to buy for me. they'll want to spend their income on other things as well. And this is where the aggregate demand externality comes in. Because whether people have enough money to buy my shoes depends on how much money they have from other sources. And if they don't have enough money from other sources,

like their income from their jobs or whatever, then they're not going to want to buy my shoes, or at least not many people will. And therefore, I'll go out of business. But this happens for every firm. It's not just shoes. It's for, like, restaurant meals or it's for cars, or it's for producing, it's for like modern supermarket.

or making clothes or electronics or whatever it is, right? Each firm individually can't survive because basically everyone's too poor to buy its output. And conversely, and this is where the externality comes in, when the firm sets up, it provides, it gives people income, which then gives them ability to demand or buy goods and services from other businesses, not just its business. If they only bought from you, say if the employees at the shoe factory only bought things at the shoe factory, then this wouldn't be an issue because there would be no externality. It would be internalized.

but that's obviously not how it works. If people get income from you, then they want to spend that elsewhere. They don't just want to spend it on you. And therefore, there's an external effect here that will actually make it easier for other businesses to exist because now some people have got incomes. So effectively, there's two equilibrium here, one in which there are lots of different businesses or making lots of different things and people learning high incomes and spending that money on different businesses.

Or there's another equilibrium in which there are hardly any businesses and hardly anyone's making money and therefore hardly anybody spending money. and there's no patronage for the businesses. This is where the idea of an aggregate demand externality comes in. It's basically the idea, again, that you need to have a customer to make a business worthwhile, and it wouldn't be an issue if people only bought things at the firms that they worked at, but that's not how it works. So you need a kind of coordination of different firms or kind of going into business at once,

or at least in similar times. And so this is thought to be a big problem in very poor countries where there's just not the market, there's not the local market for high amounts of production in many things. And so this will lead to the existence of a dual economy. You can sell small numbers of things, but the market is only small because people are poor. So basically it goes like this. A small market leads to people being poor because you can't sell very many of things. And because you can't sell very many, you're not very efficient, which also contributes to being poor, but it also means that you can't establish more efficient techniques which require you

to scale up and be able to sell more. But people can't buy that because they're poor. So there's a vicious cycle to it in which you can't scale up to more efficient technologies because the demand doesn't exist because people are poor. So I think that's a very important insight, the notion of aggregate demand externalities, and we'll revisit this idea again in the future. Another model relates to the rate of growth of technology in countries, in follower countries compared to the world leader countries. So the world leader is like the country that has the most developed technology.

So for about the last century, this has been the United States. Again, it doesn't mean that they're like the most advanced in everything. But it means they're one of the richest countries in the world and they have kind of the most advanced of a range of technologies. And so you kind of compare your growth rates to them. In the 19th century, this was Britain, in the 20th century and the early 21st century, this is the United States. So anyway, what we do is compare how a country is growing to the growth rate of the world leader. The important point here is that in order to experience the maximum rates of growth, a country needs to balance out the amount of activity or effort it puts into imitation compared to innovation.

imitation means basically taking something that someone else has already done and doing it in your country, doing it at home. So often this means hiring experts from overseas and then buying machinery or blueprints or whatever else and bringing them to your country and building the factories or implementing the techniques, building the roads, whatever else it is, implementing all of the modern advances and getting them to work in your case. So there's always going to be adaptions that need to be made so that they are. fit your particular circumstances, and that's where the imitation comes in. So there's some amount of work that needs to be done there, and that is important. But it's different to true innovation, which requires developing, you know, distinctly new things that other people haven't done before. And imitation mostly happens in poor countries, because they're just looking to the rich countries,

say, well, we just kind of need to do what they've done, but just implemented in our local circumstances. Whereas innovation mostly happens in the rich countries, because they don't have anyone to look forward to, for the most part. they need to develop new things. So that's why we see. I mean, this is precisely what we see. In China, a lot of, in recent decades, imitation, basically doing things that the West had been doing, a lot of industry, a lot of manufacturing, a lot of infrastructure. As you get closer to the frontier, however, imitation works less and less well because basically there's fewer low-hanging fruit for you to pick

that you've imitated most of the easy things, and it gets harder and harder. You get slower and slower returns to that investment. And therefore, you need to rely more and more on innovation, developing new things that no one else has done before. The point there is, though, that the institutions that support these activities are not necessarily the same. Some institutions may be set up for imitation and other set up for innovation. The classic example of this is central planning in the Soviet Union.

That was all set up under the days of forced industrial relation under Stalin, that I've talked about in episode two, which was all based on imitation. Look at what they're doing in the West, particularly Germany, before World War II. Take those ideas, take those factory blueprints and machinery, and just do that here. Build those big factories and industrialize, make more coal, steel, and all that sort of thing. That was all limitation. And their institutions were set up to focus on that, to focus on the types of things that Western countries were doing in sort of early to mid-20th century. The problem was that they did not have the institutional setup to foster innovation,

doing things that are new, because doing things that are new is risky and requires a lot of failure to find the things that work, whereas imitation is less risky because someone's already done it before, so you can kind of copy them. So the Soviet Union had very poor institutions and policies to foster innovation, and what happened is they gradually fell behind. They did not experience much success with electronics or the internet. Once they saw that, you know, in the 70s, they started to really get going with semiconductors and so forth in Western countries. They started to try to copy this later in the 80s, but were never very successful, because again, they were behind. They were just trying to imitate.

They didn't have the innovation coming into this. And there were some innovation in certain areas, like mostly in say rocketry, which was pushed strongly by the state, you know, for the space race and missiles and all that. But for the most part, there was not a lot of incentive in their economic system to innovate and do things more efficiently and effectively. And so, you know, you did not get consumer electronics in the survey union until, like, right near the end where they tried to imitate but never did a very good job. So the point there is that as a country develops, it needs to shift its institutions away from imitation towards innovation.

And you can construct models, which shows that if they're misbalanced, if the incentives are too much one way or too much the other way, you'll get stuck in a, you can get stuck in a non-conversions trap, which means that you never catch up to the West because you don't have enough innovation. You need to have enough innovation to gradually sort of move into that high-level technological growth pattern. Otherwise, you'll be sort of constantly always a few steps behind the West, trying to catch up, but never with enough innovation of your own to sort of fully get there, if that sort of makes sense. So I think that's a really important inside of that model. You need to be able to balance out imitation and innovation and gradually shift towards innovation as you get richer. Two more models that I want to briefly discuss before we finish today. One model is one in which governments are not able to pre-commit to moderate levels of taxation. So basically this means that there's an incentive for them to see.

say now that I will only tax moderately and therefore, you know, do make all your investments now and then we'll tax them next year, but we'll have moderate taxation. But when next year comes, they can't restrain themselves and it's in their interest to tax as much as they can. It's kind of like you say, well, you know, I'll go on a diet tomorrow and then when tomorrow comes, the incentive is just to not go on a diet. It's basically the same thing. If you don't have a way of committing, pre-committing and constraining yourself, then there's kind of no incentive to restrain when the circumstance comes along. Now, the problem with this is, of course, if you can't pre-commit,

then businessmen and other investors will know that, and therefore they won't invest today because they know that we get all taxed away tomorrow. This is generally a problem for weak states that don't have strong institutions that allow them to pre-commit. So states that have strong parliaments and elections, for example, have mechanisms that enable them to pre-commit to more moderate levels of taxation. Indeed, you might say that perhaps they don't tax enough in some cases,

but that's a whole other question. States need to have appropriate technological, institutional, and cultural factors that allow them to to credibly maintain moderate levels of taxation. Most states, most modern states have that, but not all of them do, particularly poorly functioning states in many of them are in sub-Saharan Africa, not exclusively there, though. And so this inability to commit can mean that you never see investment, basically, because people hear what you're saying, they hear that the state says, hey, you know, we're going to be fine. And this isn't just taxation, by the way. This includes things like bribery or corruption

and other stuff like this that can undermine incentives. But they'll say, you know, it's all going to be fine. Don't worry. Invest now and we'll only moderately tax you. But you know that that's just talk. And in fact, there's no incentive for them not to strip you of everything you have. And so you don't invest or invest very little or try to hide it as much as you can. And all that leads to very inefficient outcomes with not the most modern techniques being used and so forth. So that's a way that a country can be stuck in a low level, especially if in order to get the institutional and technical requirements to be able to pre-commit, like a functioning parliament and free press and communication and all that stuff, you need to have a certain level of wealth.

And this is another example of a poverty trap. It's something about poverty that prevents you from getting out of poverty. Final model is in which we have a trade-off between innovators and rent-seekers. In the previous one, we talked about the trade-off between imitation and innovation, this one is slightly different. The basic idea here is that there's different uses that talent can be put to in any country, really. You can try to produce wealth by innovating, by engaging in trade, by starting a business, by developing inventions, research, and there's many ways you can do that. Or you can try to take the wealth that others have already created by rent-seeking. So this includes things like, well, traditionally it might be like

joining a gang that actually goes around and beats people up, that's less common in developed countries now, but there are still other forms of rent-seeking. So much lobbying, political lobbying, is rent-seeking. It's trying to convince people to pass laws that hurt other people and help you. That's not creating wealth, usually. That's just moving it to you. So that's an example of rent-seeking. A lot of lawyers are engaged in rent-seeking. This is not in terms to be a criticism of that profession because lawyers are important for an efficient and modern society. But it's just a fact that a lot of what litigation does, especially in the corporate sense and with patent law and other things

is basically moving money around from one person to another. It's not creating new wealth. And that's what we call rent-seeking when you're just trying to get money at the expense of other people rather than producing wealth of benefits both people.

Organised crime is usually an example of rent-seeking, not always, depending on how you look at it, but like protection rackers are a classic example of rent-seeking. You're not providing a service usually. You're just extracting money from people. So organized crime is another one, or gangs often engage in rent-seeking in various ways. corrupt politicians engage in rent-seeking, either directly by taking bribes or indirectly by, you know, making policies for their friends.

So this is a problem in all countries, really, but especially in poorer countries. It's much worse there. So there's many forms that rent-seeking takes. Sometimes it's more obvious, sometimes it's less obvious, but it's a problem in all countries. And the issue with rent-seeking is that it doesn't create wealth. It just moves it from one place to another, generally sort of destroying someone or siphoning it off along the way as people try to protect their wealth or divert resources to preventing you from stealing it, essentially. or from transferring it over. Or as different people compete against each other

to have the best lawyers or the best lobbyists and the only people who get richer, the lawyers and the lobbyists and the countries a whole gets poorer. That's the basic idea of rent-seeking. Again, not saying that lobbying and lawyers don't play important roles, but I'm just saying this is a problem

in a lot of societies. And ultimately, what it comes down to is what are the returns to innovating compared to the returns to rent-seeking? If the returns to rent-seeking are high, then basically everyone's going to become a rent-seeking, and then everyone's trying to rent-seek off everyone else, which basically means everyone's kind of,

loosely speaking, trying to steal everyone's wealth. The issue here is that there's a sort of an equilibrium sort of trade-off aspect to it because the more rent-seekers there are, the more it pays me to do rent-seeking as well, potentially, or to put it another way, if I'm an innovator, I want to be in a place where there are lots of other innovators because there's spillover effects we've talked about before. Innovations and other fields give me ideas,

and they make the country richer, which leads to new technologies and so forth, or other people get richer, so they have more money to buy my product. Generally, innovators and businesses like having other businesses and innovators around unless they're direct competition. But usually there's spillover and beneficial effects there.

On the other hand, if you have a situation which there are lots of rent seekers, you do not want to be an innovator because all these rent-seekers are going to come in and try to take rents from you, try to take their returns to what you've produced. Likewise, you might also have to be forced into rent-seeking to defend yourself against other rent-seekers.

So the point is that there's sort of two-week liberate. There's one with lots of innovators and a few rent seekers, and there's one in which there's few innovators and lots of rent seekers. And it's thought that this might describe different countries, particularly at different times, sometimes in which a lot of people are going into rent-seeking activities, say working for the government or law or other things, or organized crime, or in other cases, more people going into trade and sitting up businesses or science or other things like that. Not saying that all government or lawyer work is bad or all trade is good,

but it's just to say that there is this difference between innovation and rent-seeking and that if the incentives are misbalanced, you can get too much rent-seeking and therefore have a significantly lower reward to activities that increase the economic pie for everyone. So these are just some examples of models that have been developed fairly recently in the last couple of decades that attempt to describe different pieces or aspects of why countries grow at different rates or why some countries get stuck at low levels. Again, they don't represent a full theoretical description or full picture. They just sort of pieces of the puzzle.

And I think you have to combine many of them to get a full picture. Also, these theories still don't really explain why particular countries succeed and others fail. Often it will be something like, well, this is a mechanism by which if you're poor, you're probably going to stay poor. That's mostly what we've talked about here. We haven't really been able to say, how do you get out of these poverty traps? And why do some countries get out of the poverty traps, but others don't? That's the real question, right? So in other words, we've got a lot of ways, looking back at all the growth theories we've discussed,

we've looked at a lot of ways in which that if you start to grow, you can grow more up to the point where you get those diminishing returns, right? However, we also looked at a bunch of ways in which you can fail to grow at all. That's the coordination failures and poverty traps. So growth theories tell us all of these things, but they don't really answer the fundamental question, why are some countries able to get on the growth train and, you know, reach the solo swan steady state and, you know, continue to grow under the spillovers in the Roma model and endogenous growth theory through improvements in technology.

Why do some countries manage to escape the poverty traps and get onto that solar swan Roma growth train? And others don't. And why do some countries fail for a long time and then succeed? Like China failed for a long time and then succeeded in the late 70s. And other countries have failed and still continue to fail. And some countries succeed for a while and then fail, or at least fail relatively. Like Mexico's growth has slowed down a lot in the last couple of decades. So these questions are not really answered by growth theories.

In the next episode, we're going to look at some of the theories that have been put forward to try to explain why some countries are able to get on the growth train and others are stuck in the poverty traps. So look forward to that. I hope you enjoyed this episode. If you'd like to leave feedback about the show or ask questions or anything, you can get in touch with me via email. My address is Fods12 at gmail.com.

That's FODDS12 at gmail.com. I also appreciate any support that people might be willing to provide the show. you can go to my website, which is Fods12.podbean.com and click on the support the show link there, and there's links to the Patreon page, and also PayPal, if you prefer to donate that way. I really appreciate everyone who's made contributions.

It helps me to be able to devote more time to the show and cover various overheads. So that's much appreciated. Once again, thanks very much for listening, and I'll talk to you next time.