"Exogenous" and "endogenous" are two words that get thrown around a lot in higher-level economics courses. Since these are real words and not specifically "economics terms", classes seldom define them for you (just like the word "autarky"!). But it can really help to know what they mean before starting some upper-level courses!

These concepts are often encountered when discussing macroeconomic models, or in econometrics contexts – like when conducting regression analysis. Put simply, "exogenous" is an adjective that refers to a variable, quantity, or phenomena that is derived outside of the current economic model or context. The Merriam-Webster dictionary defines it as follows: "caused by factors…from outside the organism or system". In economics courses, this typically means that you aren’t expected to know how the exogenous variable came to be. Its origins, and how it is determined, are not part of the analysis.

"Endogenous" is the opposite; it’s something that is defined within the current context. Merriam-Webster defines it as: "[something] caused by factors inside the organism or system". This means that, given the current economic model, you are expected to be able to describe the origins and determination of the endogenous variable, unlike the exogenous case.

These concepts are fairly intuitive once grasped, and often form the basis of more advanced assumptions in econometrics and economic model contexts. To illustrate this further, let’s use a simple market equilibrium example.

Market equilibrium example

Suppose that we’ve been given information about a market, and we want to find out what the market equilibrium will be. Suppose that the supply curve is defined by QS = 2p, and the demand curve is defined by Q_D = 12 - p. Then, equilibrium is where Q_S = Q_D, or where 2p = 12 - p. Solving, we arrive at an equilibrium price of p = 4, and therefore an equilibrium quantity of q = 8.

In this example, we had two equations to describe the market’s equilibrium — the supply side and the demand side. With these two equations, we were easily able to solve for the resulting equilibrium. The equilibrium price was partly determined by the equilibrium quantity, and vice versa.

Which parts of this problem — which variables, and which key concepts — were endogenous, and which were exogenous?

The equilibrium price and the equilibrium quantity were both determined by the equations we had, and we calculated them as part of our main analysis. Therefore, these variables were endogenous.

Conversely, the constant in the demand equation (12, in this case) was a number that was simply given to us. We don’t know how that number came to be there, nor were we concerned with it. It was exogenous.

But there’s more. In order to find the price and quantity, we needed to use two equations that were defined for us: the supply curve and the demand curve. How did we know what those were? How did they come to be in such simple linear forms?

We have no idea! In this case, both the supply curve and the demand curve were also exogenously given. It wasn’t part of our analysis to determine them, and we took them as given.

This simple example points to a useful rule of thumb when trying to decide if something is exogenous or endogenous. Chances are, if a variable, equation, or fact is simply given and isn’t “supposed” to be questioned, it’s probably an exogenous variable. It’s needed to conduct our economic analysis, but how it came to be is not relevant.

Another way to think about this is to consider which variables are held constant; these are probably exogenous. In our example, the number “12” was just a constant and was exogenous, while both q and p were not.

Conversely, the variables that are given equations or relationships to other variables in the system are probably endogenous. If we changed the price (not the equilibrium price) in the previous example, the quantity would have needed to change, too, and vice versa; but the constant in the demand equation (12) wouldn’t be affected by just p or q changing. That, too, hints that “12” is exogenous and both p and q are endogenous.

This is likely a good enough explanation for most introductory economics courses, but we can dive deeper. Let’s take the ubiquitous Cobb-Douglas production function as an example.

Exogeneity and endogeneity in Cobb-Douglas

The basic form of the Cobb-Douglas production function is as follows:

Y = AL^𝛼K^𝛽

where Y is economic output, A represents technological progress, L stands for labor, K stands for capital, and 𝛼 and 𝛽 are constants. In this case, A, 𝛼, and 𝛽 are often considered exogenous; they are constants whose values are usually given to students, and they aren’t affected by other variables changing.

But, this equation defines Y as a function of other economic variables, including the factor inputs of labor, capital, and technological progress. Thus, we can’t define Y without first knowing the values of all other variables. Clearly, in this case Y is endogenous. Any change in another variable will change Y, and further, the point of the Cobb-Douglas function is to see how economic output Y changes when other factors change.

Sometimes, it may be unclear whether a variable is endogenous or exogenous. That’s because sometimes, economists can choose! Consider both L and K in the Cobb-Douglas equation above. If values for these variables are given to you on an economics test, then they’d most likely be exogenous for the purposes of the question.

However, macroeconomists may use the Cobb-Douglas framework to examine how economic growth changes over time as a response to growth in L and K. To do this, they’d collect a time series of the values of Y, L, and K for many years, and use a computer to fit an equation to them and determine their relationship.

In this case, L and K can be considered endogenous! In a time series, previous values of L and K will affect future values. That’s because the population grows at a rate defined by the size of the population itself, and because the economy doesn’t destroy all of its factor inputs and restart from scratch every year. So, L and K must be determined from their own starting values and their growth rates…but the starting values and growth rates can be thought of as exogenous, since we may not know exactly how they came to be.

Exogeneity and endogeneity in regression analysis

These terms appear very often in econometrics contexts, particularly with regression analysis. In regression analysis, there is typically one dependent variable (usually called the y-variable) that the analysis attempts to explain with a set of independent variables (usually called the x-variables).

This language already hints at which variables are endogenous and exogenous in this setting. The dependent or y-variable is usually endogenous, and very often the purpose of the regression is to study how that endogenous variable is affected by, or dependent on, economic forces in the real world.

Those economic forces are often measured by collecting data, which then form the independent or x-variables. These are often considered exogenous; the regression model is not concerned with how those variables came to be, but rather with how they affect the y-variable that the model is attempting to explain.

For example, consider a regression model that attempts to explain house prices (the dependent or y-variable) using the size of the house in square meters, the distance to the nearest grocery store, and the quality of the school district as independent variables. In this model, the house price is endogenously determined, while the house’s size, proximity to groceries, and location in a good school district are exogenous factors that the price itself probably does not influence (said differently, those factors likely influence the house’s price rather than vice versa).

This is a general rule of thumb, but there are a wide variety of regression models and a wide variety of economic variables, and not all of them follow this pattern.

Be careful about assuming exogeneity or endogeneity!

For example, suppose you’ve constructed a simple regression model to see how various economic factors affect annual GDP. You collect data on inflation, unemployment, and whether or not there was a recession to use in this model.

The dependent variable is GDP, which you might consider endogenous, since you’re constructing a model to see how other factors affect it. In this case, you might consider inflation, unemployment, and recession to be exogenous; they’ve been collected to explain how GDP changes from year to year.

But there’s a problem with simply applying these labels, because GDP itself can affect all of these other variables! If an economy has a higher GDP in a certain year, it’s much less likely to be in a recession, it’s more likely to experience some inflationary pressure, and it’s more likely to see a reduction in unemployment over the period.

This is a thorny case where it’s very difficult to say if any of the variables in the model are truly exogenous. Since all of them can affect each other, they may all be endogenous after all. Advanced regression techniques — such as finding suitable proxy variables — can be employed to test these theories and help the model work.

Fortunately for this article, the moral of the story is already clear: it can be very helpful to be aware of the difference between “exogenous” and “endogenous”, to begin understanding other economic relationships. This will be particularly important for higher-level economics courses.

After all, your test scores will be endogenously determined by your own understanding (and diligence in studying)!