About the Econometric Institute

The Econometric Institute at Erasmus University Rotterdam has a strong research tradition in econometrics, statistics, data science and operations research. Founded in 1956 by Henri Theil in cooperation with Jan Tinbergen, it is the oldest research institute in the field of econometrics in the world. 

Today, the Econometric Institute offers several Bachelor and Master programmes in Econometrics and a substantial number of well-known econometricians have their roots in Rotterdam. The institute is part of an active research community within Erasmus School of Economics, and participates in the research networks and graduate programmes of the Tinbergen Institute, Erasmus Research Institute of Management (ERIM), and LNMB (National network for Operations Research).

The research of the Econometric Institute covers three research areas and six application areas:

Research Goal

The overarching goal of the research conducted within the Econometric Institute is to advance methodological knowledge related to econometrics, statistics, data science and operations research. It is our mission to help address societal, economic and business-oriented challenges that require complex data analysis and/or optimization. We aim to be internationally recognized as one of the best places to study econometrics and to do research in this area. We aim to have our methods used by other academics and in industry to have major impact. An overview of the institute’s publications can be found on the Pure publications page.

Research Themes

Econometrics

In the Econometrics research area, we focus on advanced statistical methods and techniques that are useful in economic context. This research pushes the state of the art in econometrics to provide economic agents, including policymakers, firms and investors, with quantitative tools to make the best possible decisions. Within Econometrics we distinguish three research themes: Time series econometrics, causal inference, and panel data econometrics.

Time Series Econometrics

This research theme concerns the development and application of methods and techniques that accommodate the key features in the dynamic behavior of (macro-)economic and financial variables: long-term trends, seasonality, time-varying volatility, nonlinearity, and structural breaks. Special attention is devoted to time-varying parameter models and factor models. Forecasting is an important use of time series models; where different types of forecasts may be considered, including point forecasts, interval forecasts, probability forecasts, and density forecasts. Research on forecasting involves both the construction, as well as the evaluation of the quality of such forecasts. Finally, Bayesian time series analysis and in particular structural VAR models are considered.

Causal inference 

Causal inference amounts to estimating the independent effect of a phenomenon on a variable of interest. Research focuses on estimating (heterogeneous) treatment effects using synthetic control, difference-in-difference methods, and instrumental variable approaches.

Panel data Econometrics

We consider panel data models for a large number of individuals and/or a large number of time periods. New methods are developed to estimate network effects and quantiles using panel data models. Further research concerns pooling and clustering in linear and nonlinear panel data models, including limited dependent variable models. 

Operations Research

In the Operations Research area, we are at the European forefront of the developments of Operations Research models and algorithms with applications in transportation and logistics, and in energy and sustainability. Next to developing new methodological advances, we focus on having a real impact on society.  The main research theme of the Operations Research group concerns optimization.

Optimization

We aim at developing analytical methods to compute (near)-optimal solutions for decision-making problems. These methods include deterministic optimization methods such as linear programming, non-linear programming and integer programming, as well as stochastic optimization methods such as stochastic programming and Markov decision processes. Both exact methods and heuristic approaches are developed.

Data Science & Statistics

This research area aims at developing and analyzing statistical tools and machine learning methods for empirical data analysis. The research of the Data Science & Statistics area can be summarized into two main research themes.

Machine Learning

This research focuses on developing supervised and unsupervised machine learning methods that have applications in the field of social science, in particular economics and business. Methods include regularized estimation techniques (such as Lasso, Elastic Net, and through Bayesian priors), clustering, neural networks, boosting, autoencoders, and regression trees. Also, natural language processing methods are developed to extract information from unstructured text. The final topic focuses on high-dimensional data visualization. 

Mathematical Statistics

The research focuses on deriving statistical properties of estimators and testing procedures. There is a special focus on modeling extreme events and robust inference. Deriving and analysing the properties of e-values is another field of interest.

Application areas

Energy & Sustainability

We provide analytical methods to improve decision-making in energy and sustainability. Research topics are energy systems and sustainable health. Renewable energies have different characteristics than traditional fossil fuels due to the uncertainty in supply (wind and solar). Therefore, new forecasting models and stochastic optimization techniques are required to solve these challenges.

Transportation & Logistics

The research group applies analytical methods to improve decision-making in transportation and logistics. We contribute to efficient supply chain management by studying, production planning, inventory management, and last-mile logistics. Transportation focuses on freight transportation and public transport optimization. 

Climate

One part of this research theme focuses on measuring and modeling extreme weather events, which is of increasing interest and importance in the current era of climate change. Another part of this research theme examines the consequences of climate change on different areas in economics and finance; for example, the impact on risk exposures of long-term asset allocation strategies of institutional investors such as pension funds.

Policy evaluation

This research theme concerns evaluation of policies and treatments. Research focuses on analyzing the effects of macro-economic shocks, announcements and policies and macroeconomic forecasting. Additionally, treatment effects, like the effect of schooling, are estimated using non-experimental data.

Finance

Quantitative tools are essential in the context of important financial decision-making problems. These include empirical asset pricing; portfolio optimization; risk management and hedging; and derivatives pricing. We develop and evaluate novel statistical tools that can be used for these purposes, including models and algorithms for predicting returns and volatility; methods for assessing market risk and default risk; and for asset portfolio construction. 

Analytics for Business

Quantitative support is crucial for many business functions. We contribute to such support by developing econometric and data science techniques for, for example, demand modeling, online marketing, recommendation systems, and consumer or investor sentiment analysis.