The Science of Everything Podcast - Episode 36: Consumer Choice Theory

Episode Date: August 12, 2012

A discussion of the theory of consumer choice in economics, including an outline of consumer preferences and the axiomatic assumptions economists make about them, a discussion of the realism of these assumptions, and an explanation of how preferences are used to derive consumer utility functions and demand curves. These basic concepts are then applied to understand consumer behaviour in an analysis of substitution and income effects, complementary and substitute goods, and elasticity of demand. Recommended prerequisites include Episode 12: The Price System and Episode 16: Profits and Competition.

you're listening to The Science of Everything podcast, episode 36, consumer choice theory. And I'm your host, James Fodor. In this episode, we're going to look at the area of microeconomics called Consumer Choice, or just Consumer Theory. This field establishes a mathematical model of consumer preferences for goods and services and what people want to buy and how much they want to buy and things like that, and then puts these things together to try and work out individual and collective consumer demand for various goods and services and how they're going to do.

these are changed by changes in income and changes in prices and things like that. So although it might not sound like the most interesting topic in the world, it is very important to economic analysis and sort of underpins a lot of the other more applied fields of economics that I'd like to talk about in future episodes. So that's why I wanted to do an episode on this so we could get that basic foundations. The sort of sister field to consumer theory is producer theory which looks at how businesses produce goods and services, how their cost structures will affect the profits that they can make and the amount of goods they sell, and also how the different levels of competition and industry structure in different markets will affect

the outcomes there. We'll do that in a future episode. For this one, we're going to look at consumer theory and preferences and demand and so on. Recommended prerequisite episodes for this episode are episode 12 on the price system and episode 16 profits. and competition. First of all, we need to explain what we mean by the idea of preferences. I should point out that the stuff that I'm going to do talking about is mostly mathematical in nature, but I'm going to try and present it in a way that is non-mathematical, obviously, because of the nature of the podcast. So I might be a little bit vague in parts, and some of the concepts are more mathematical than anything else. So I'll just try and present

it in a conceptual way that can allay to get the broad idea of what the theory says and what kind of things it predicts. Okay, so what do we mean when we talk about preferences? Before we can start to talk about the demand for goods and services or how much consumers will buy and so on, we need to have an idea of what the basis is for making those decisions. And preferences is a way of understanding that. Preference series basically just asserts that, well, suppose consumers have preferences over different bundles of goods. So there are many thousands of different types of goods and services that are sold or available to be sold in the world. And you can imagine consuming these in various

different combinations, three apples and two bananas, or four apples and five bananas, five apples, a car, and a book, you know, and so on and so forth, with varying different quantities and amounts and qualities of all these different goods. We could put these in a whole range of different bundles and then ask a consumer, or ask many consumers, which of these bundles would you prefer? Would you prefer three apples and two bananas or two apples and three bananas or whatever. And generally we think, and indeed people do this in experiments, they'll be able to rank the bundles that are offered to them in order of preference. And so we say that if a consumer ranks one bundle higher than another, then they prefer that bundle of goods to another. Makes sense. Now, the reason we do it this way

is because what we'd really like to be able to do is work out exactly how much a consumer like to give a bundle of goods compared to another bundle of goods. This is the idea of utility, which will come back to in a second. That's like how much benefit you get out of something, or how happy it makes you. But this approach was sort of taken earlier on in the history of economics, but it was sort of abandoned because you can't measure utility in any meaningful way, even theoretically. You can't measure the subjective satisfaction that someone gets from something.

So all you can do is work out on a behavioral basis, even if this is sort of a hypothetical behavior, Which would they prefer? Would they prefer bundle A or bundle B or bundle C of goods? Which collection of goods and services would they prefer over another one? You won't be able to tell how much they prefer it, but you'll just know that it's preferred to another. Okay, so having this concept of preferences, this can be formulated like in mathematical set theory and so on. We won't worry about that. But in order to develop the model, what's necessary is to make a few assumptions about the nature of these preferences,

because just having an idea of purely of preferences doesn't necessarily tell you very much. the preferences could be in any sort at all. So economists make a few assumptions, which are borne out to varying degrees by the empirical evidence, and I'll talk about that in a second. But they make a few assumptions about the nature of these preferences, which helps them to make more specific predictions from the model of consumer choice that we're building. Okay, so the first assumption they make about preferences is that preferences are complete,

which simply means that any two bundles can be compared, and the consumer can choose which one they prefer, or say that they don't care. That that means they're indifferent between the two bundles. Now, this seems like a fairly innocuous assumption. There doesn't seem to be any reason why this would be false. I'm not aware of any experimental results or even theoretical reasons why this wouldn't be the case. If you present someone with two bundles, they'll always either tell you,

I prefer bundle A, I prefer bundle B, or I don't care. There's no inability to make a ranking there. It might be difficult to decide sometimes, but as long as the decision can be made, then complete, is satisfied. Okay, so that one's fairly innocuous. Assumption two is continuity. Now, this is mostly for mathematical convenience, but it sort of means that there are no sudden reversals and preferences, so that similar bundles will be treated similarly. So that would be sort of like, if you like three bananas and two apples, then all of a sudden when we went to four bananas and two apples, you wouldn't dramatically lower that in your preferences. That's not a very

satisfactory explanation of what continuity means. As I said, it's mostly a mathematical assumption and is fairly innocuous in practice. It just says that similar bundles will be sort of similar and won't be dramatically all over the place. Maybe you could think of examples where that wouldn't hold, but generally we expect it to hold, so we don't worry about it too much.

So, we've got completeness and continuity. The third assumption is transitivity. This is a word you may have heard before because it occurs in a number of context. Transitivity, or transitive preferences, means that ordering of bundles or ordering of the preferences is consistent. So, to give an example, this means if you have transitive preferences, then if you prefer apples to oranges, and if you also prefer bananas to apples, then you must prefer bananas to oranges.

So I'll say that again. It's kind of like saying if one is greater than two, sorry, if three is greater than two and two is greater than one, then three must also be greater than one. Now, that seems obvious, and it is obvious when we're talking about numbers, but when we're talking about sets or just groups of things, it's not necessarily the case. or it doesn't have to be the case at least. But in consumer theory, we assume that it is the case. We assume transitivity. Now, this is a very important assumption, because if you have intransitive preferences, it means at least in theory,

a consumer could be vulnerable to exploitation in a phenomenon called the money pump. So this basically means if you have intransitive preferences, say someone gave you, say you had an apple. Well, actually, say you just had a few dollars and you were willing to buy an apple for it because you like apples. So the person sells you an apple. But then they produce an orange, and they say, I'll sell you this orange in exchange for your

apple plus another dollar. And you might be willing to do that because you like oranges more than apples. So that means given a bundle of one orange and a bundle of one apple, you prefer the bundle of one orange to the bundle of one apple. You prefer to have an orange to an apple, at least when you only have one of each, which is all that's relevant in this case. Okay, so you're willing to buy the apple originally, then you're willing to give the apple away and an extra dollar in exchange for the orange, because you like the orange more than

apple. Nothing wrong with that. But then, suppose the person pulls out a banana and says, would you be willing to give me a dollar and give up your orange in exchange for this banana? And again, you say yes, because you prefer bananas to oranges. You like oranges better than apples, which is why you're willing to pay that extra dollar and give up your apple to get the orange, but you like bananas even more than oranges. So you're willing to give up the orange and an additional dollar in order to get the banana. And again, that's fine. There's nothing wrong with that. However, now suppose the guy pulls out the apple again and says, will you give up your banana and an additional dollar

in order to exchange the banana for this apple? Now, if you had intransitive preferences, you could prefer apples to bananas, even though you preferred oranges to apples and bananas to oranges, which means their preferences are in a sort of a loop. It's sort of like, oranges is greater than apples, but bananas is greater than oranges, but apples is greater than bananas. So it goes round in a loop. Which means you would, and if your preferences were like that,

if they were intransitive in this way, you would be willing to give up your banana and an additional dollar in exchange for the apple that you started with. But then the person could just say, well, here's an orange, and will you give me up your apple and another dollar in exchange for the orange? And you say, well, yes, I like oranges better than apples, so I'll do that.

Then he pulls out the banana. Would you be willing to give up the orange and a dollar in exchange for this banana? Well, yes, I'll be willing to do that because I like bananas better than oranges. Then he pulls out the apple again, and, well, would you be willing to give up your banana and an additional dollar in exchange for this apple?

Well, yes, you say, I would do that because I like apples better than bananas. And so it goes on. So this is the idea of a money pump that they could just keep selling you, like going through the cycle of things and selling you the same thing, extracting money from you. Now, in the real world, obviously a situation that simple wouldn't arise, but you could still imagine or conceive of more complicated ways of effectively doing the same thing, rebuying and selling or involving consumers in transactions like that

that effectively just extract money from them without really providing anything. taking advantage of their in transitive preferences. So the fact that that could occur is a very strong reason why economists suspect that transitivity holds at least most of the time.

So in consumer choice theory we assume that it always holds, although there are variants where you can relax that assumption, but assuming in the basic theory that it always holds doesn't mean in the real world that it actually does hold.

So this theory of transitive preferences has been tested on a number of occasions. And I'll post some of the links up to these papers on the website or on the Facebook or something, because this is a very difficult assumption to test. The details are very tricky and technical, but the basic idea is that transitivity is a mathematical or like a logical relationship,

but in practice you can't test a logical relationship, all you can do is test statistical relationships like how often is an intransitive choice made in a given experimental setup, and how often would we expect such a choice to be made just by chance or just by like a mistake being made, because people make mistakes in the real world. And there's a lot of different ways of measuring how many, like how intransitive preferences are, because when you talk about loops, you can have loops of different sizes, and anyway, it gets very complicated. So there's no agreed-upon way of measuring transitivity very well.

Some experiments have produced results which claim to be clear violations of transitive preferences. Others have not. my personal opinion looking at the research is that intransitive preferences probably do occur under some situations and you know depending on a whole bunch of factors like the environment that someone's in and the mood there and things like that that can affect choices and you can refer back to some of our cognitive bias episodes about things like that so yeah probably intransitive preferences do occur but are rare

especially in the real world when making important decisions because if you did have intransitive preferences you would quick you would make a large number of irrational decisions and if they were prevalent enough in the world, we would see very clear consequences of them. And that's more of a theoretical reason for expecting it, and the empirical evidence seems to support the idea that intransitive preferences are rare at best. We don't find them very easily and repeatedly in studies. At best, we sort of, maybe that's intransitive. Okay, so that's our third assumption of transitivity, which seems like it holds reasonably well most of the time. There are two more assumptions. The

assumption of monotonicity, which sounds really scary. It just means more is better. This assumption just says that if one bundle contains more of everything of all goods and services than another bundle, then you will always prefer that to the other bundle. So, you know, if one bundle has one apple, one banana, and one orange, and a second bundle has two of all of those things, then you'll always prefer the second bundle to the first bundle. Once again, this is an assumption that almost certainly does not hold in the rural world, because the only time it would hold is if you had free disposal, which means you can just throw away goods, get rid of them, if you didn't want them, without any cost. In practice, that's obviously not the case. However, for most instances,

more is better, or monotonicity probably holds. Generally, if you can get more of everything, you're going to prefer that to less of everything. So that's another one where probably doesn't always hold, well, it almost certainly doesn't always hold, but probably holds enough to be useful. Because remember, this theory of preferences is designed to be useful and predict behavior and model situations in the most common types of market transactions, and non-market transactions too, but it doesn't need to be robust to every single little potential exception or strange circumstance that could occur. So we're just trying to get something useful here, not necessarily universally true. So that was the fourth assumption, monotonicity, or more is better. The final one

is convexity. Now this assumption is not essential, but it's it's usually added on because it makes the math easy, basically. Convexity is another technical assumption. Basically, it means diminishing marginal benefit. Now, I should explain what that means. The marginal benefit that you get from something is the extra benefit that you get from consuming or buying or having one additional unit of the substance.

So I may have mentioned this in a previous episode, but I'll just quickly go through it again. So, for example, if you only had access to a single glass of water all day, the marginal benefit that you would get from that would be very high, because you need to drink that water in order to stay alive. Second glass of water would be very high as well, slightly less because you've already had one glass to drink, but you'd still need a second glass to drink.

By the time you've got many glasses of water, you're going to start using the... You don't need to drink it anymore because you've had enough to drink. You'd probably use it for cooking or for washing your hands or something like that, which is a pretty important use, so your benefit's going to be pretty high, but not as high as for the first couple of glasses which you needed to drink. So your marginal benefit,

the benefit that you get from those extra few glasses is going down. By the time you have heaps and heaps of water, you've already used, you already have plenty for drinking, plenty for cooking, plenty for washing, you're probably using it for things like watering your lawn or washing your car or something like that. That's not that important. And in fact, when you have, like most people in the developed world, when you have heaps of water access, it's not really, the marginal glass of water that you have is probably not used for washing your car. It's probably used for being slightly less careful with the hose or the bucket when you are washing your car. But anyway, so you can see how the extra benefit that you get, goes down as you get more and more of something. In this example it does because we're using the case of water. So this is what diminishing marginal benefit means. As you get more and more of something, the extra benefit that you get from one more unit goes down. Another simple example of this is, you know, food like donuts, for example. If you like donuts, the first donut's great. The seventh donut, nah, not so good. And the 14th donut, you're probably throwing up. So you've got negative benefit there. So anyway, back to our assumption of convexity. Convexity basically means

a diminishing marginal benefit, which means that as you get more and more of just one thing, the benefit goes down. And that means that, in reference to our bundles of goods, consumers generally prefer mixed bundles or balanced bundles of goods to extreme bundles. So this would mean, for example, that if you had convex preferences, you would prefer four apples and four bananas, or you would likely prefer four apples and four bananas to nine apples and one banana, for example. Once again, this assumption certainly doesn't hold in all cases, and it's not one of the crucial ones. The first three that I mentioned are the really crucial ones.

Completeness, continuity and transitivity. However, convexity makes the math easier, as I said, and it probably holds a lot of the time. Probably, we generally expect people to prefer balanced bundles over extreme ones. And, in fact, we can look at household data and see that people spread their expenses over a wide range of different areas, and they don't just spend inordinate amounts of money on one over the other.

So that seems to be borne out in the data. Okay, so those are our five assumptions about preferences. We'll just go through them quickly. Completeness, which means you can rank any bundles. You can compare them and say which one I like or which one I don't like compared to the other. Continuity, which just means there are no sudden reversals and preferences. So similar bundles are treated similarly. Both of those are pretty innocuous.

Transitivity, which is pretty controversial, but probably holds in most cases. Transitivity says that you don't have sort of circular preferences, that the preferences are consistent. Fourth assumption, monotonicity, or more, is better, just means that if one bundle contains more of everything than another bundle, then you'll always prefer that one to the other bundle, and that probably holds in most cases. And finally, convexity, which means you have diminishing marginal benefit of goods, and once again, that probably holds in most cases. Okay, so those are our assumptions.

So what's the good of all of these technical assumptions and all of our hypothetical preferences and ranking bundles and all that? I mean, what does it have to do with anything? Well, you can use that preference theory to make simple predictions of consumer behavior just by itself, but economists have extended that theory to be more useful. And one important way they've done that is by the idea of a utility function. So what is this? This sounds very mathematical and confusing. Remember the concept of utility that I mentioned before.

Utility refers to, well, it's not a very well-defined concept. It doesn't really need to be. It just refers to the benefit or the benefit that you get from something or the happiness you get from it or the pleasure, anything like that. Any sort of positive psychological connotation you want to throw on it works well enough. I like to think of it as the benefit that you get from something, but it's a subjective psychological sense. And so you can't measure utility.

And economists don't really bother with that. Although there is some work in like neuro-economics, neuro-economics, which sort of uses MRI in regards functional magnetic resonance imaging while engaging in economic experiments to try and tease out, like, which regions of the brain and neurotransmit is involved in that sort of thing. We'll probably do that in a future episode, but that aside, economists don't worry about measuring utility. What they do is they use it as a tool. So it turns out that if you have preferences as described by our previous assumptions that I just mentioned, if you have those, there's a proof, which we, of course won't bother with,

that you can represent those preference relations or the choices resulting from that by a mathematical formula called a utility function. Function is just like a rule or an equation which tells you how much benefit that you'll get or how much utility that you get from a given bundle. So essentially it'll say like if I have five oranges, three apples and two bananas, I'll get such and such amount of utility, but if I have three apples, ten bananas and eighteen oranges, I'll get such and such and such other amount of utility. And as long as you have those assumptions about preferences that I mentioned before, you can do that.

It'll work. Just any old sort of preferences cannot be described by a utility function, but those that satisfy our assumptions can be described by such a function. Okay, so why would we want to do this? Well, the reason is that if you have a function, a utility function, for an individual or for an organization or something like that, you can use that to predict its behavior. So you can say that if a consumer has these preferences and also our assumptions about those preferences hold, then we can, as I said, according to this proof, we can describe their

behavior as if they were maximizing their utility function. By maximizing, we simply mean making it as big as possible, or getting as much utility as they could as they can from a given amount of, from a given amount of income. Now, we want to say maximizing the utility function, that's not quite correct, because the maximization is constrained. What this means is that there's a limit to how much utility we can get, and that limit's generally going to be our income, or possibly our wealth or something akin to that, but generally you think of it as income, So if you had an infinite amount of money, you could theoretically have an infinite amount of utility, because you could just buy everything. But in practice, of course, we can't do that.

And so when an economist is trying to use a utility function, they'll say, how much income do we have to work with, or how much income do we sort of think is reasonable in this case? Or depending on the circumstances, so they'll take some level of income, and they'll say, subject to this amount of income, you know, given a hundred bucks, what is the most benefit that you could get from that? What is the highest utility you could get? and we can maximize the utility function and we can do various mathematical manipulations to work that out, and we can also work out then how much they would spend on each of the goods that we consider, that we're considering. So, you know, as long as you have the utility function to find, you could work out that given $100 they'd spend X amount on apples, X amount, Y amount on bananas, and Z amount on oranges or whatever. And if their income change, you could work out how much their demands for each of those goods would change as well.

So that's very useful. Now remember, the actual utility as such, that the number that we get out from that doesn't matter. All that matters is that we can use the utility function to make predictions about demand and substitution and things like that, which I'll come to shortly. The other thing that's important about utility functions is that we need to be able to specify

what the utility function looks like, because different utility functions represent different underlying preferences, and different people have different preferences. And so that's the tricky part, or one of the really tricky parts about consumer theory, knowing what the utility function looks like. You could try and measure that experimentally, but that's very difficult. Usually, economists will abstract a bit, a lot, actually,

and just use general forms of utility functions that they think are reasonable. But I'll talk more about that in a little bit. Okay, so before we advance to the next part about demand curves, let's just take a quick recap of what we've done. What we're trying to do is we're trying to build a theory or a model explaining how consumers make choices about what to buy, given certain incomes and how that changes with prices and income and so forth. To do that, we assumed that consumers had preferences over bundles of goods,

bundled just different groups of goods and services that they could buy. We assumed that these preferences were subject to five assumptions, which we mentioned before, and it turns out that if consumers do have preferences, and if these preferences do abide by these five assumptions, then we can describe consumer behavior as if the consumer behavior, as if the consumers were maximizing utility functions subject to their budget constraint, meaning that given a certain amount of money, consumers will buy goods and services or split up their money on spending different things

in accordance with how our utility function would predict. Now, this model is often criticized on the basis that it's unrealistic. It is unrealistic. Any model is unrealistic. That's the whole point of a model, because if a model was perfectly realistic, it wouldn't be a model anymore. it would be the reality, which is not very useful because we're trying to understand the reality, not just sort of duplicate it. The other main criticism is that consumers don't do anything like this. Consumers don't know what a utility function is, and they don't think of themselves as maximizing utility functions and so forth. And that's certainly true. Consumers mostly have no idea what a utility function would be.

But the point is that doesn't matter. Economists don't say that people do maximize utility functions. The economists wouldn't say that there is a utility function inside someone's head, and they are consciously or even unconsciously maximizing it when they make decisions. All economists would say is that the way they behave is consistent with the way someone would behave if they did maximize a utility function. So it's an as-if argument. They're behaving as if they maximize utility function, even though that's not actually how they're making decisions.

Even that sort of reduced claim is also false, because if you look at the field of behavioral economics, you can see many exceptions to utility maximizing behavior. However, there are a few extra assumptions which really should have mentioned before that this utility theory requires. And that's basically perfect rationality and perfect information. Perfect information about you know what you're consuming and how much it's going to cost and you know how much income you have and things like that.

Which is obviously unrealistic, but it's used as a benchmark measure and to make the first case analysis simple. And then when we want to make more realistic, we can introduce. uncertainties. So in perfect information, the other one is perfect rationality. That's not really an extra assumption. That just sort of follows from our assumptions about preferences. Perfect rationality would fail if the assumptions we made about preferences, you know, the five assumptions did not perfectly hold, or hold under all circumstances, or if there are other influences on preferences that we're entering there, which most certainly is the case in the real

world. So this utility maximizing behavior is not going to perfectly or necessarily even very accurately represent reality. The proof is in the pudding really about how accurate are the predictions it makes and how useful is it for modeling various outcomes. That will look at in future episodes when we try and apply this sort of theory to real-world cases, like, for example, in labour economics, health economics and stuff like that. And on the whole, I would say it's fairly useful. Basically, because although people don't have perfect information and although people are not perfectly rational, they do have some information and they are somewhat rational, and so they make decisions that are somewhat and reasonably consistent with those of utility maximization theory.

So, you know, when the price of something goes down and other things are being equal, people buy more of that and less of something else. When something gets more difficult, people, and the benefits of doing it roughly the same, people will do less of that thing. You know, these are the things we would expect from utility maximization theory. Okay, so that's where we are at the moment. I now want to move on to talk about demand curves and elasticity and substitution and income effects, which is where all this preferences and utility stuff has been leading, and this is a really interesting part of it, I would say. Now, a demand curve just tells us how much of a given type of good the consumer is willing to buy

depending on prices and income. You could have a demand curve for a single individual, or you could have one for the market as a whole. We'll just think about one for an individual person for the moment. Now, you can get these demand curves, that is you can derive them mathematically from a utility function, turn comes from underlying preferences. So demand curves fall directly out of the assumptions we make about preferences. That is, if a consumer has preferences and the preferences fulfill the assumptions that we make, they'll have a demand curve as well. And the demand curve will satisfy certain properties that make it relatively easy, at least, to deal with. So that's

the good thing about demand curve. Basically, it tells us some basic things. Like generally, if the price of something goes up, consumers will buy less of it. They'll buy something else instead. Or if their income goes up, consumers generally buy more of something. The down the cost of the downside to demand curves is that we can have individual consumer demand curves, that's no real problem. But when we try and have an overall market demand curve, it becomes problematic. In other words, you can't aggregate individual demands up to form a market demand curve. They're very complicated reasons why you can't do this. The basic idea is that as prices change, some consumers gain in income and others lose income, because prices relate to individuals' income, because wages and

interest rates and stuff like that are prices, but they also relate to income. So as prices change, income changes. And so you've sort of got two things changing at once, and that screws up the whole demand curve thing, which assumes that the demand is independent of your income level. The demand will change when you change income, but it won't change when just prices change. So that means we can sort of talk meaningfully about, well, what is the demand response, the quantity demanded response of a consumer to a change in price? But if prices affect income, then you've got too many things happening at once and you can't meaningfully talk about it. So aggregate demand curves, that is, like, you add everyone's demand up for apples or everyone's demand up for cars or something like that.

Don't work so well in the same way. Economists do use these. In fact, that's sort of mainly what they use, but you have to make a few extra assumptions, manifestly unrealistic ones, if you want to derive them directly from preferences. A better solution there is simply to say that we'll assume that the overall demand curve like slopes downwards, which essentially means that when the price goes up, people buy less, which holds in pretty much all circle. So it seems a reasonable enough assumption to make. So that's demand go, we can get that from preferences. What else can we get from this theory of preferences?

Substitution and income effects. Now, these are a really important concept. Substitution and income effects are two different ways in which demand, that is the quantity that you demand, the amount of something you want to buy, two different ways in which the demand responds to changes in prices. Let's imagine we have a consumer who likes to buy oranges, and the price of oranges goes up, a reasonable enough amount so that it makes some difference. Now, as a result of that, two things have happened.

One is that the price of oranges is higher compared to other things. So I guess I should have added, we'll assume that everything else stays the same. The only thing that's changed is the price of oranges goes up. So the price of apples is the same, the price of bananas is the same, the price of bananas is the same, etc. When the price of oranges goes up, oranges are now relatively more expensive compared to other goods. So this will lead to what's called a substitution effect. The consumer will substitute away from oranges to.

towards buying other things. The substitution effect is always negative, which, and this is sort of what's called the law of demand, it means that when a good gets relatively more expensive compared to other goods and your income stays the same, you'll always buy less of that good. Or technically, you might still buy the same, but you'll never buy more of that good. This comes out of the basic assumptions about preferences and really makes sense and is born out by empirical evidence. Basically, if something gets more expensive and your income stays the same, there's no reason at all you would buy more of it. You may still buy the same, you may buy a bit less, you may buy a lot less, but you're not going to buy more of it.

Because if you wanted to buy more, why wouldn't you have bought more in the first place when it was cheaper compared to other things? So that's the substitution effect. The income effect is the second thing that happens when the price of oranges goes up. When the price of oranges goes up, the consumer is now poorer than they were before, as long as they are at least spending some positive amount on oranges. If they spend zero on oranges, then that's not going to make any difference. But the reason they're poor is because their real income,

that is their income taking into account the prices that they spend for things, has gone down because income, like in dollar terms, they own $10,000 a year, and the price of most things is the same, but the price of oranges has gone up a bit. So their real income adjusted for prices has gone down just a little bit. And if it's the price of cars

or the price of housing that goes up, then maybe their real income goes down by a lot, because they spend a lot on those things. But you probably don't spend that much on oranges, so a rise in the price of oranges is not going to reduce your income by much, but it will reduce it by a bit. And that has an effect.

It has an effect on many things, but it will have an effect on the demand for oranges, which is what we're interested in in this case. However, unlike the substitution effect, the income effect can be positive or negative, so it can work in either direction, just depending on the underlying preferences, basically. So this means that the individual might buy more oranges

or fewer oranges as a result of their income going down. Their real income going down. Remember, I said income is constant. What I mean by that is nominal income, income measured in dollar terms, is constant, but real income adjusted for the actual price you pay for things has gone down because prices have gone up, or at least the price of oranges has gone up in this case. So just a quick recap, so far we've got the substitution effect, which is always negative.

That means when the price goes up, you always buy less of something. That's simply because of the change in the ratio of prices of different goods. And we've also got the income effect can be positive or negative, and that is due to the change in your price-adjusted income that results from changes in prices of goods, because when the prices of good change, your real income changes and therefore you may buy more or less of the good in question. So just to motivate these concepts a bit, let me introduce you to a sort of an interesting paradox that this theory is useful for explaining, which is the concept of a given good. One example of this is occurred in the Irish Potato Famine, or at least this is given as an example. It's not 100%. I'm not half a percent

sure if it is a true example, but it's a good one to illustrate the point anyway. And that is that during the Irish potato famine, when the price of potatoes went up dramatically, that reduced Irish Peasants' incomes to such an extent that they actually bought more potatoes. That is, the price of potatoes went up, but the demand for potatoes went up regardless, which is the opposite of what you would normal expect to happen. So according to sort of traditional ideas that the demand cove slopes downwards and when the price goes up, you always buy less, this would be impossible. However, we can understand this, apparently paradoxical situation using the concepts of the income and substitution effects that we've just outlined. When the income effect is positive, this is called

a normal good. That just means that when you get richer, when you get richer, when you're your income increases, you buy more of the good. It's called a normal good because most goods are like that. You get richer, you want more of it. But some goods are called inferior goods. When you get richer, you want less of it. Maybe cheap wines or low-quality meats would be an example of these things, or just cheap housing. It's unlikely that whole classes of goods would be inferior, but specific types of them may be. So like the cheap versions would be inferior goods. You want to move up to better quality things so you actually buy less of some things when your income goes up. Those are called inferior goods. Now, be careful, though. This is just the income effect.

This is the effect purely of changes in income on your demand for some good. This is not the substitution effect. The substitution effect is always negative. So when the relative price of something goes up, you always, always buy less of it, just considering that effect. When your income goes up, you may buy more or less of something depending upon whether it's a normal good or an inferior good. Now, in the case of a Giffin good, which I mentioned a second ago, that is where the income effect is negative. so you buy less of something as you get richer, or equivalently you buy more of it as you get poorer.

So that has to happen. But also the income effect has to be bigger than the substitution effect, which means if the price of something goes up, your income goes down, which increases your demand for the good, that's the inferior goods part. However, the price of it's gone up, so the substitution effect is still negative.

The substitution effect is always negative. However, if the positive income effect, because it's an inferior good, exceeds the negative substitution effect, then the overall demand, change in demand, will be positive. That is, you'll buy more of something when the price goes up, even though that's normally the opposite of what happens. That's called a given good, and as I said,

it's speculated that that's what happened during the Irish potato famine. The price of potatoes went up because of the famine, sorry, because of the pestilence that was affecting the food crops, the significantly reduced incomes of many of the Irish peasants, and because potatoes were an inferior good, and because the effect was so large, their income effect was negative, and so they bought more of the potatoes. Now, that was offset somewhat by the negative substitution effect, but not enough. So the income effect won out, and overall the demand for potatoes went up.

Now, of course, if that happens, the price is also going to go up, and that will just reinforce the cycle. So that's one of the reasons why it was such a problem. When you add up the substitution and income effects, that produces what's called the total effect. Sometimes substitution and income can work in opposite directions, sometimes they work in the same direction. So that I would put as an example of a very powerful use of consumer theory, because otherwise a giff and good would just be completely unable to understand

how the demand for a good could possibly go up when its price went up. That should never happen. It contradicts the law of demand. Well, with this utility theory, we can explain exactly how it happens and even predict in circumstances when it would happen. That is when you have a single good whose price goes up that is very important to income, and that is also an inferior good. That could lead to a given good situation like the Irish Chattah famine. Another concept that's very important is that of elasticity of demand. So previously we just looked at substitution and income effects, and we're now looking at elasticity

of demand. This word elasticity sounds a bit, you know, technical again, but just think of it as being like how elastic something is, how stretchy it is, because that's where the word comes from. We know that when the price goes down, the price of some good goes down, pretty much always the demand goes up, or the quantity demanded goes up, I should say. And similarly, when the price goes up, the quantity demand goes down. Giff and goods excluded. Those are very rare anyway. So that's always the case.

The question is, how much does it go up or down? Demand might go up. Quantity demanded might increase by a little bit or by a medium amount or by a lot when price changes. And the theory itself doesn't tell us how much it would be. We have to measure that empirically, and we can do this by measuring elasticity of demand. So basically what we do... Elasticity demand is just a ratio. It says like when you have a given, when price goes up or down, but in this case we'll think up.

Price goes up by 1%. How much does demand change? If demand also changes by, if quantity demanded goes down by 1%, remember we would expect it to go down if price goes up, then that's called unit elastic. It sort of balances out. It's just sort of average. The change in demand is similar in size to the, well, it's exactly the same in size in this case, to the proportional change in price. If demand goes down by a lot more, like 2% or 3%, then we call that. elastic demand, which means that demand responded a lot to the change in price. Price changed by a bit,

but demand responded a lot. If the response in demand is small, like half a percent or 0.1 percent in response to a 1 percent changing price, we call that inelastic. It's not very stretchy. Demand doesn't respond very much to a given changing price. Now generally, goods that we think of as necessities or essential basic things have low elasticity of demand. That is, they are inelastic. The reason for this is that generally they're so cheap and also so necessary that people don't really care how much they cost. Electricity, petrol or fuel, basic foods, all fit into this category. So the demand doesn't change very much when price changes. You think about something like table salt. If the price of table salt doubled, how much less of it would you buy? Well, you'd probably buy a little bit less

of it, but not very much, and so that would be inelastic demand. On average, all goods and services have unit elastic demand, because the average of the average of the, response has to equal one, because when you spend less on some good, you have to spend more on another good. So the elastics and inelastics have to balance out in the end of the day. So sort of the average elasticity is one. Things that tend to have elastic demand, are sort of luxuries or very expensive items, things like cars and houses and overseas holidays, for examples, because those are things that you can postpone or put off or not buying completely if prices go up or your income goes down or things like that. So those are things

to which people tend to respond a lot to changes in price. However, we shouldn't think of elasticity as sort of intrinsic to the good in question. It's going to depend upon people's incomes, the social and political and economic situation, and it's also going to depend upon the amount of time that has passed. Over time, it becomes easier and easier to substitute one good for another. So substitutes are just goods that can be used, that can be swapped out for another. So a good example of that is, say, gas and electricity. You can use gas as a source of energy, or a source of cooking or a source of electricity generation, whatever, or you can use, well, I guess coal, say, coal and then that generates electricity.

Coal for electricity production or just electricity for, say, your stove or something like that, there's substitutes for each other. You can use one or the other, but you don't need both. If the price of one goes up, you switch over to the other, and so you substitute one for the other. Now, it might be difficult to make substitutes. So another example of a substitute would be instead of driving to work, you take the train, or vice versa. That might be difficult because you've got to figure out where the train leaves, and you've got to figure out a way of getting there, maybe you even have to move house. that could be tricky. So that's why we expect elasticity is to increase over time. That is,

people respond more to price changes as more time passes, basically because they have more time to change their behavior. A really good example of this is petrol or car oil. Now, people will say things like, non-economists, I should say, would say things like, well, it doesn't matter if the price of the price of oil goes up. People need to drive, so they'll just buy the same amount. This is incorrect. Remember, the substitution effect is always negative. When the price of something goes up, the demand always goes down. This always applies unless it's a given good, and oil is not a given good. We've measured this, it's not a given good. Very, very few things are. So the question is just how much of a reduction in demand will there be in response to a rise in the price of oil?

And the answer is, the size in the reduction will increase over time. So in the very short run, you know, just a couple of days after the price of oil goes up, there's not so much you can do about it. You can maybe be a bit more careful, try and drive it more efficient speed. try and avoid driving if you can, if you can walk to a house that's five minutes away instead of driving. Maybe you'll do that because it's a bit more expensive. But probably you can't reduce your demand too much. Over a bit of longer term, maybe you could get your car serviced more regularly, or you could even buy a more fuel-efficient car,

try and maybe change the route that you drive to work or something like that. Various other things you could do to try and reduce your fuel consumption. In the very long run, you might even move house to live closer to your job, or something like that, in order to reduce your fuel consumption. and whether you do that depends upon how much of a rise in the fuel price there are and a whole bunch of other factors. But everyone doesn't have to do that. As long as a few people do that, there'll still be an overall response in demand. And that's a key concept in economics, the sort of marginal response. It doesn't have to be that everyone changes their behaviour.

Even if only a few people do, that will change the overall demand for oil, which will therefore be represented by the response in demand. So the point, though, is that the amount of flexibility you have increases over time and so that the elasticity also tend to increase over time. What we've been talking about so far in terms of elasticity is more technically called own price elasticity of demand. That's when the price of oranges goes up, how much does your demand for oranges go down?

There's also such a thing called cross-price elasticity of demand. This refers to one good price having an effect on the quantity demanded of a different good. That is, if the price of oranges goes up, what happens to your demand for bananas? Or for eggs or for whatever else? Now, there's sort of three things that cross-price elasticity can be.

It can be positive, it can be negative, or it can be zero, essentially. If it's zero, it means the two goods are not related. They just have nothing to do with each other. So my demand for, I don't know, physics textbooks really has nothing to do with the price of oranges. I don't care what the price of oranges are when I'm buying physics textbooks. And I do buy physics textbooks, by the way. So my cross-price elasticity of the demand would be zero. However, the goods can also be compliments.

That is, if you have one good or if you want one good, you're more likely to want the other. So a great example of this is renting a DVD. what do you want? You want snacks, particularly popcorn or something like that. So those are compliments that go together. A car and petrol to fuel the car go together. Compliments are goods that, sort of, when you have one, your demand for the other one increases. And so, compliments tend to have negative cross-price in elasticity, which means that if the price of one goes up, you'll tend to buy less of both of them. Because say we're talking about movie tickets and popcorn prices, for example, and let's pretend these are set completely independently.

which is probably false, but we'll just assume that. So if movie ticket prices go up, that means we go to the movies less often. What does that mean? Well, it also means we buy less popcorn. So our demand for popcorn also goes down alongside our demand for movie tickets. These are compliments, so they move together. That's a negative cross-price elasticity. A positive cross-price elasticity occurs when goods are substitutes, which I mentioned before.

This means that you have one or the other, but you don't want both. Or in other words, if you have one, you're less likely to want the other. So this is the opposite of compliments. Compliments you have one, you're more likely to want the other. If you have a movie ticket, you're more likely to want popcorn than if you didn't have one because they go together. Examples of substitutes, as I mentioned before, gas and coal, for example. You don't need both of them.

You need one or the other. Another example would be a train ticket to a certain place or renting a car to go to that other place. You don't need the train ticket and the rented car. One or the other is enough. So in that case, when the price of, say, the train ticket goes up, your demand for the train obviously goes down because it's more expensive, but your demand for renting the car will go up now, because people are moving away from taking the train to taking

cars instead. This is an example of sort of responsiveness to changes in price. The more substitutes there are for a good, the more elastic, the price, the own price, response of that will be. So one reason oil has a relatively low elasticity of demand, this is own price we're talking about again, is because there aren't too many really good substitutes. However, breakfast cereals, for example, probably have a relatively high elasticity of demand. Actually, they probably still have a low elasticity of demand because they're pretty cheap. But just looking at the number of substitutes that are available, any given brand probably has a fairly high elasticity of demand. Because if that brand becomes too expensive, you'll just buy a different brand, which is pretty much exactly the same.

So the number of substitutes strongly affects the elasticity of demand. Finally, there's one other important type of elasticity, which is the income elasticity of demand. So previously I've been talking about price elasticity of demand, which is price changes, how does your demand change? Income elasticity of demand is your income changes, but price doesn't. How does your demand change? And if it's a normal good, your income elasticity of demand will be positive because you buy more of something when your income grows up. If it's an inferior good, you'll buy less of it, and so your income elasticity of demand will be negative. And the size, the absolute size of your elasticity of demand depends on how sensitive you are to changes in income.

When someone gets a lot richer, they probably don't much change their demand for toilet paper, but they probably do substantially increase their demand for caviar, for example. So their income elasticity for both of those, toilet paper and caviar, would both be positive, let's assume. But the income elasticity of demand for toilet paper is probably pretty low, that for caviar probably pretty high. Okay, so that's the main topics I wanted to cover in this episode. I hope it wasn't too dry. As I said, it's mostly setting up for future episodes.

when we need to apply these concepts and so it's useful to have them. So just a very quick recap, we talked about how consumers have preferences, the five assumptions that we place on those preferences, how we can use these assumptions about preferences to construct utility functions and describe consumer behavior as maximizing that utility function to do the best you can, given a restricted amount of resources or income that you have access to. We talked about how that you can then use this utility maximization to derive demand curves, how much of a certain goods or a bunch of goods consumers will pay,

how much they'll buy, depending on how much income they have and what prices are. We talked about how these levels of demand will change when prices and income change. These are our substitution and income effects. And we also talk about the relative sizes of those changes, which is our elasticity of demand and income elasticity of demand. So I hope you enjoyed this episode. Remember that the podcast is on Facebook now. So go to The Science of Everything podcast on Facebook.

Just search us there and like the podcast. You can also send me an email if you have suggestions or feedback. praise, criticisms, whatever. Fods12 at gmail.com. Thanks for listening, and I'll talk to you next time.