If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

Generalized linear consumption function

Generalizing a linear consumption function as a function of aggregate income. Created by Sal Khan.

Want to join the conversation?

  • blobby green style avatar for user pearcenesbitt
    If Csub1 times T is taxes and Y is income at the vertical axis won't income be zero and thus taxes zero so the addition of the minus c sub 1 to the C naught is superfluous?
    (25 votes)
    Default Khan Academy avatar avatar for user
    • mr pants teal style avatar for user Wrath Of Academy
      No, this is a really bizarre model. He's assuming that taxes are being taken independent of the size of the economy. Even if there was no economic activity in the country, the government still somehow collected T taxes.

      You always have to take these models for what they are. They're just silly little ways of helping you think about bigger problems. Don't take the models too seriously. Sal says so in several of these, but it's worth repeating.
      (6 votes)
  • blobby green style avatar for user Jordi H Urquidez
    Can autonomous consumption ever be 0?
    (9 votes)
    Default Khan Academy avatar avatar for user
    • primosaur seedling style avatar for user Cameron
      In practice this is impossible. Autonomous consumption is consumption expenditure that occurs when income levels are zero. Because humans have necessities, we will need to consume whether or not we have income. We may borrow money just to buy necessities. So consumption of food, water, and other necessities can be seen as autonomous consumption. If your autonomous consumption is zero, you are dead
      (29 votes)
  • blobby green style avatar for user Liam
    is autonomous spending inclusive of mortgage repayments, bills and so on. Essentially is disposable income all income left over after all financial obligations have been met
    (4 votes)
    Default Khan Academy avatar avatar for user
    • starky ultimate style avatar for user Geoff Ball
      Good question. Disposable income is personal income after taxes and government transfers, but not all financial obligations. That is, disposable income does not mean money you're free to spend as you wish (since your bank won't really like it if you skip out on your mortgage), but rather the money you're left with after the government is satisfied.
      (6 votes)
  • leaf green style avatar for user Anupam Misra
    when Income is zero, will not T be zero as well? In that case how the slope will look like?
    (6 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user saurav agrawal
    How can income be the GDP?
    (5 votes)
    Default Khan Academy avatar avatar for user
  • hopper happy style avatar for user Sharpy
    wouldn't you have to subtract Co (autonomous consumption) from Y to make it disposable income after your autonomous consumption so the equation would be C=Co+C1(Y-Co-T)?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user Wincy Heung
    How does both share of population and entire of population are credit constrained affect the slop? Thanks
    (2 votes)
    Default Khan Academy avatar avatar for user
  • duskpin ultimate style avatar for user DS27Lee
    I was taking AP Macroeconomics in another class and they made a point you guys didn't quite address.

    At when you start graphing you give the equation
    C[sub1] * Y + c[sub0] - C[sub1] * T

    you don't mention that since C[sub1] is not constant and is the slope, if C[sub1] changes, wouldn't it change the Tax amount as well? Thus, shifting the line a little up and down since b(y-int.) is autonomous consumption minus MPC*Taxes?

    Ex. that AP Macro showed in the class I was taking

    When MPS(or C[sub1]) increases, the slope of the line becomes steeper but does not exceed one. But the line shifts downward a little since the value of C[sub1] * Taxes has become greater and the Y-intercept of C[sub0] - C[sub1] * T has become less.

    Normal(C[sub0]) - increased (C[sub1] * T)

    But the shifting concept was not introduced in this video or any of the other ones. So is this concept applied in the Macroeconomics world?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user tskachuwa
    how does exchange and interest rates get affected by changes inflation
    (1 vote)
    Default Khan Academy avatar avatar for user
  • mr pink red style avatar for user Peijia
    For aggregate income, should't it be consisted of disposable income, taxes and also transfers? Why transfer isn't included?
    (1 vote)
    Default Khan Academy avatar avatar for user

Video transcript

Male: In the last video, we began our exploration of what a consumption function is. It's a fairly straightforward idea. It's a function that describes how aggregate income can drive aggregate consumption. We started with a fairly simple model of this, a fairly simple consumption function. It was a linear one. You had some base level of consumption, regardless of aggregate income, and then you had some level of consumption that was essentially induced by having some disposable income. When we plotted this linear model, we got a line. We got a line right over here. I pointed out in the last video this does not have to be the only way that a consumption function can be described. You might use some fancier mathematical tools. Maybe you can construct a consumption function. You have an argument. You would argue that the marginal propensity to consume is higher at lower levels of disposable income and that it kind of tapers out as disposable income, as aggregate disposable income goes up. You might think that maybe you should have a fancier consumption function that when you graph it would look like this and then you would have to use things fancier than just what we used right over here. What I want to do in this video is focus more on a linear model. The reason why I'm going to focus on a linear model is because, one, it's simpler. It'll be easier to manipulate. It's also the model that tends to be used right when people are starting to digest things like consumption functions and building on them to learn about things like, and we'll do this in a few videos, the Keynesian Cross. What I'm going to do is, I'm going to do two things. I'm going to generalize this linear consumption function, and I'm going to make it a function not just of disposal income, not just of aggregate disposable income, which is what we did in the last video, but as a function of income, of aggregate income. Then we will plot that generalized one based on the variables. It's really going to be the same thing. We're just not going to use these numbers. We're going to use variables in their place. Let's give ourselves a linear consumption function. We can say that aggregate consumption where we're going to have some base level of consumption no matter what, even if people have no aggregate income, they need to survive. They need food on the table. Maybe they'll have to dig in savings somehow to do it. So, some base level of consumption. I'll call that lower case c sub zero. Or lowercase c with a subscript of zero right over there. That's the base level of aggregate consumption or it's sometimes referred to as autonomous consumption. This is autonomous consumption because people will do it on their own, or in aggregate they will do it on their own, even if they have no aggregate income. Then we will have the part that is due, directly due, to having some aggregate income. We call that the induced consumption, because you can view it as being induced by having some aggregate income. Above and beyond what the base level of consumption, people are going to consume some fraction of their disposable income. So we'll say disposable income. They're not going to consume all of their disposable income. They might save some of it. So they're going to consume the fraction that's essentially their marginal propensity to consume. This right over here, I'll do that in this orange color. Marginal propensity to consume. Hopefully this makes intuitive sense. This says, look, if this was 100, people are going to consume 100 no matter what, 100 billion whatever your unit of currency is. Now, if their marginal propensity to consume is, let's say, it is 1/3. You have now above and beyond this people have disposable income of let's say 900, this is saying that they want to consume 1/3 of that disposable income they're getting. That is, if you give them 900 of extra disposable income, they're propensity to consume that incremental income, they're going to consume 1/3 of it. So this would be 1/3, so it would be 900. Let me give an example. If you had a situation, you could have a situation, where c-nought is equal to 100. If you have disposable income is equal to 900, and c1 is equal to 1/3, or we could say 0.333 repeating forever, c1 is 1/3. Then this makes sense. On their own people would consume this much, but now they have this disposable income. Their marginal propensity to consume if you give them 900 extra of income, they're going to consume 1/3 of that. So then you're going to have, your consumption is going to be equal to, for this case right over here, your consumption is going to be 100 plus 1/3 times 900. So your consumption in this situation, your induced consumption, 1/3 times 900, would be 300, maybe it's in billions of dollars, 300 billion dollars. Then your autonomous consumption would be 100. They would add up to 400. Once again, this is autonomous and this is induced. Autonomous, this right over here is induced consumption. Now, I did write it in general terms. I'm using variables here instead of, or constants, really instead of using the numbers we saw in the last example. But I also said that I would express aggregate consumption as a function not just of disposable income but of aggregate income; not just of aggregate disposable income but aggregate income. The relationship is fairly simple between disposable income and overall income. We saw over here, in aggregate, you have income, but the government in most modern economies takes some fraction of that out for taxes. What's left over is disposable income. Just a reminder, income in aggregate, aggregate income is the same thing as aggregate expenditures, which is the same thing as aggregate output. This right over here is GDP. So this right over here is, let me do this in a color, I've used almost all my colors. This is equal to GDP. Disposable income is essentially GDP, or you could say aggregate income, minus taxes. I'm going to do the taxes in a different color. Minus taxes. So we can express disposable income as aggregate income, this right over here is the same thing as aggregate income minus taxes. We could rewrite our whole thing over again. Aggregate consumption is equal to autonomous consumption plus the marginal propensity to consume times aggregate income, which is the same thing as GDP, times aggregate income minus taxes. We fully generalized our consumption function and now we've written it as a function of aggregate income, not just aggregate disposable income. To make you comfortable that this is still a line if we were to plot it as a function of aggregate income instead of disposable income, let me manipulate this thing a little bit. We could distribute c1, which is our marginal propensity to consume, and we get aggregate consumption is equal to autonomous consumption and then we're going to distribute this, plus c, so we're going to multiply it times both of these terms, plus our marginal propensity to consume times aggregate income, and then minus our marginal propensity to consume times our taxes. Since we want it as a function of aggregate income, everything else here is really a constant. We're assuming that those aren't going to change. Those are constant variables. What we could do is we could rewrite this in a form that you're probably familiar with. Back in algebra class you probably remember you can write it in the form y=mx+b where x is the independent variable, y is the dependent variable. If you were to plot this, on the horizontal axis is your x axis, your vertical axis is your y axis. This right over here would have a y intercept, or your vertical axis intercept of b, right over there. Then it would be a line with slope m. If you were to take your rise divided by your run, or how much you move up when you move to the right a certain amount, that gives you your m. Slope is equal to m. The same analogy is here. We can rewrite this in that form, where our dependent variable is no longer y. Our dependent variable is aggregate consumption. Our independent variable is not x, it is aggregate income. So let's write it in that form. We can write it as dependent variable, c, which we'll plot on the vertical axis, is equal to the marginal propensity to consume times aggregate income, I'll do that purple color, times aggregate income, plus autonomous consumption, minus marginal propensity to consume times taxes. It looks all complicated, but you just have to realize that this part right over here, this is all a constant. It is analogous to the b if you were to write things in kind of traditional slope intercept form right over here. When we plot the line, if you have no aggregate income, this is what your consumption is going to be. Let me draw that. Once again, our dependent variable is aggregate consumption. Our independent variable in this is no longer disposable income like we did in the last video. It is now aggregate income. If there's no aggregate income, this is the independent variable right over here, if there's no aggregate income, then your consumption is just going to be this value right over here. So your consumption is just going to be that value right over there, which is c-nought minus c1 times t. Then as you have larger values of aggregate income, c1, that fraction of it, is what's going to contribute to the induced consumption. What you essentially have is this is the slope of our line, this right over here is our slope. Just to kind of draw the analogy, if you were to say y is equal to mx plus b. Actually, maybe I'll write it like this. If you were to write c is equal to m ... and I don't want to confuse you if this m and b seem completely foreign. It comes from kind of a traditional algebra grounding in slope and y intercept. If I were to say c is equal to my plus b, this is the slope. This is our vertical or our dependent variable intercept right over here. That's where we intercept the dependent variable axis. And this is our slope. It's our marginal propensity to consume. Our line will look something like this, where the slope is equal to the marginal propensity to consume, which is equal to c1. If people all of a sudden are more likely to spend a larger fraction of their income, then the marginal propensity to consume would be higher and our slope would be higher. We would have a line that looks like that. We always assume that the marginal propensity to consume will be less than 1. So we'll never have a slope of 1. We'll also never have a negative slope because we assume that this is positive. If people are more likely to save than consume when they have extra income, then this line might look something like that. It might have a lower slope.