EFA (20): Ceteris paribus

You do not have to be a scholar of Latin to understand economics. There are, however, a number of Latin terms that are commonly used in the subject.

The most important Latin phrase is probably "ceteris paribus". This is the subject of the 20th item in our Economics for Amateurs (EFA) series. You can find the series here each Monday.

Even Maynard, my economist parrot — who you can see here, thinking about the best solution to the current banking crisis — knows how to say ceteris paribus. And he know what is means, too.

Literally, ceteris paribus means "with other things the same", although the normal English version is "all other things being equal".

This concept is important anywhere one wants to identify and isolate the causal relationship between one variable and another.

Let's take a simple example from economics: if we want to know what effect an increase in the price of oil will have on the demand for petrol, then we need to hold all other variables constant (income levels, population levels, tastes, etc.).

In theory, we can be pretty sure that an increase in the price of oil will, ceteris paribus, lead to a fall in the demand for petrol. But we may not observe this correlation in practice because, in the real world, all other things do not remain equal.

For example, if incomes are rising worldwide, as they have been over the past decade, particularly in China, we may observe rising oil prices and rising demand for petrol at the same time.

This might seem like a paradox at first sight. Surely, the higher price of oil can't be the reason why people are buying more petrol. And indeed it isn't. They are buying more because they are richer and this positive "income effect" is overriding the negative effect of price on demand.

In laboratory experiments in other subjects, it is often possible to hold all other factors constant in order to identify causal relationships between specific variables. In the real world of economics, this is not the case.

As a result, economists have to use sophisticated statistical techniques to try to identify specific causal relationships. The use of such techniques is known as "econometrics".