# What is Econometrics?

Econometrics is a discipline which combines mathematics with the study of economics to analyze information and work with mathematical models to explore scenarios and theoretical situations. The study of econometrics tends to be especially strong on statistics, with statistics being a powerful tool when it comes to analysis, and it requires comfort in the fields of both economics and mathematics. This academic pursuit was created in the 1930s, when numerous mathematicians started to explore mathematical relationships in the economic world. Today, it is offered at numerous educational institutions, and leading researchers in this field can be found all over the world.

A number of people use econometrics, including mathematicians and economists who work on a purely theoretical level. Banks and governments also utilize this tool to explore potential scenarios and ways in which shifts in wages, interest rates, and so forth would influence the health of an economy. By using tested models based on historical events, econometricians can come up with theories about ways in which various events will test or change the economy.

In order for econometrics to be effective, the data being used must be accurate, as must the model. Using a model with clearly stated assumptions is critical, as is having a sound reason for those assumptions. For example, a model might assume that a reduction in interest rates leads to a rise in lending, basing this assumption on analysis of lending rates after interest rate changes. Econometrics must also take cultural factors into account, as economies around the world behave very differently in part because of cultural differences.

The language of econometrics rapidly gets very complex, and it can be difficult for lay people to understand. While specialists in this field do occasionally interact with the general public, they usually work behind the scenes. Econometrics is at work in news reports about predicted changes in the economy, for example, with reporters relying on statistical projections to look at ways in which the economy may shift and evolve.

Statistics can be a slippery field of mathematics, because it is easy to manipulate statistics to push for a desired outcome. Economic analysts try to be careful about this, as they do not want to influence their results. The use of peer review is also heavily emphasized in economic analysis and econometrics, to ensure that published works are vetted by the econometrics community in general. This increases the validity of the information while promoting cooperation and dialog.

## Discussion Comments

@nony - It’s interesting that you mention gold. I’ve actually seen gold prices plunge in value at various intervals. I don’t know of a single econometric analysis that could predict such as change; the assumption seems to be that gold rises constantly with inflation fears.

I do think mathematical models serve a useful purpose however, because certain correlations tend to remain stable over time. For example, the article talks about reduced interest rates leading to increased lending.

I think that tends to be true. What is not predictable is consumer behavior, however. For example, banks may be willing to lend, but are businesses willing to borrow? If businesses feel that the current economic climate is suffering under a weight of government regulations, they may not even borrow the available money for expansion.

That’s where the models break down. I don’t think there is a single mathematical variable for “government regulation.”

Econometrics sounds like a useful tool but I would take some of its assumptions – and especially its projections – with the proverbial grain of salt.

I once read one guy’s mathematical model of the ratio between the price of gold and the real inflation rate. He claimed that for every $100 increase in the price of gold, inflation would increase by a certain rate.

Gold went through the roof in the decades following, but the inflation rate stabilized at a certain level. My opinion is that some of these models may be accurate to a certain point (like for the first few hundred dollar price increases in this example) but then tend to break down after that.

The reason is that there are other factors that influence inflation and interest rates more powerfully, in my opinion.

## Post your comments