In economics, **econometrics** is the use of mathematical and statistical methods to produce models to test the validity of assumptions about economic policy.

Econometrics is now a branch of economic science; but to know it thoroughly, one must keep in mind that in its time it was also a movement advocating a new direction of research in economics and that at present it also constitutes a professional qualification. The name itself, still little used at the turn of the century, began to spread after the Econometric Society was created in 1930. Its founders – a number of European and American scholars – intended to promote the use of quantitative and mathematical methods capable of giving economic theory the rigor, precision and objectivity it lacked and making it more usable by public authorities.

The movement, in which a number of future Nobel laureates in economics took an active part, developed rapidly, despite certain ideological biases on both the right and the left: while around 1950 there were no more than a few dozen participants in the Society’s European congresses, by about 1970 there were about five hundred. The Society’s journal, “Econometrica,” founded in 1933, became in the postwar period one of the most prestigious among economists around the world.Forty years ago the term ‘econometrician’ denoted both a mathematical economist and a researcher devoted to estimating relationships or testing hypotheses by means of statistical observations; after all, the two qualifications were often associated in the same person. Today, the former meaning has all but disappeared: practicing econometricians–in university research centers, government departments, and large companies–are those who deal with statistical induction on economic data, or who work on quantitative models that reproduce economic phenomena, such as the macroeconomic evolution of a country.

As a discipline, econometrics is first and foremost a method for the inductive treatment of observations, analogous to that adopted in the natural sciences for using experimental data. Second, it is a complex of techniques developed to apply this method within the framework of specifications that are often peculiar to economic phenomena and from non-experimental data. Finally, it is a complex of results obtained by induction as certain practices followed to predict the effects of changes (suffered or intended) in the environment in which economic activity takes place. Therefore, mathematical economics, which is concerned with the deductive determination of the properties of theoretical models used by economists, should be considered outside econometrics today.

## Purposes and techniques

The main purposes of econometrics are to give empirical content to economic theory, and to subject the latter to statistical testing. For the most part, econometrics is based on results from classical statistics. Of these, the econometrics scholar’s most important statistical tool (or at least the one most relevant in practice) is probably linear regression, as well as linear model estimation by the method of least squares. More sophisticated applications resort to the methods of maximum likelihood and method of moments.

A growing number of researchers, on the other hand, make use of tools and methods peculiar to Bayesian statistics. The latter approach, characterized by a subjective interpretation of probability, seems quite promising in the field of econometrics applied to finance, or financial econometrics.

In applications, econometrics studies can be divided into two broad categories: time-series analysis and analysis on cross sectional (or cross sectional) data; recently, the increased availability of data has made possible analyses based on panel data, which incorporate both time and cross sectional dimensions.

Although historically econometrics studies developed from the goal of studying models proposed in the field of macroeconomics, in recent times the need to empirically ground economic arguments has led to a broad development of econometrics studies, applied to different branches of economics. Alongside the studies of finance mentioned above, one cannot fail to mention the remarkable growth of microeconometrics, which has led to remarkable achievements in the field of labor economics and the study of market regulation.

## Differences between econometrics and statistics

Econometrics is now considered a separate discipline from statistics because it focuses on the collection and analysis of nonexperimental data, that is, data not collected through experiments. In fact, the use of experiments and laboratory environments for data collection in the social sciences is a rare phenomenon (it would usually have too high a cost or require hard moral choices).[1] Over time, econometrics has borrowed much from statistics. For example, the method of multiple regressions is used in both disciplines, but its purpose and interpretation can vary widely. In addition, econometricians have developed new techniques to handle the complexity of economic data and to test the predictions of economic theories.