Skip to main content

Identification and Structural Inference of Dynamic Causal Effects: Theory and Applications

Article Category

Article available in the folowing languages:

New macroeconomic theories redefine inflation

Deeper research in structural inference of dynamic causal effects has helped to examine inflation from a new perspective.

Industrial Technologies

The study of macroeconomics is being put into question after the recent global economic crisis which continues to impact our economies. Tackling macroeconomic issues such as inflation can benefit from new research in structural inference – i.e. causal inference relying on underlying subject matter theory or structure to identify exogenous variation in observational data. The EU-funded ISIDCE (Identification and structural inference of dynamic causal effects: Theory and applications) project investigated new methods of structural inference and their application to macroeconomic theory. Through structural inference, it examined whether expectations about future inflation are an important determinant of current inflation. This can support policymakers in improving their management of expectations in order to meet their policy objective of controlling inflation. The idea that past inflation affects current inflation through its impact on future inflation expectations is considered a 'weak' instrument in studying inflation. In this context, the project team improved upon existing methods of inference that are robust to weak instruments. It worked on new methods that exploit previously unexplored variation in the data and applied those methods to current open questions in macroeconomics. Specifically, the team has successfully improved the scope of existing methods of inference and has published the results in Econometrica, a leading peer-reviewed academic journal of economics. It also articulated a new method of causal inference that can be applied to inference based on time series data. The method stipulates using variation induced by structural change as instruments to identify causal effects that are stable over time. To illustrate, ISIDCE developed a new method of identifying causal effects from non-experimental data and applied the results to two important macroeconomic models. This method employs variation induced by structural change to identify causal effects that are stable over time. Its results were published in two papers that deal with the estimation of the New Keynesian Phillips curve (supply side) and analysis of the Euler equation model (demand side). In addition, ISIDCE analysed the New Keynesian Phillips curve and proposed new insights to understand its complexity, along with ideas for future research. These results were disseminated at key conferences worldwide and through a wealth of publications. The outcomes have also attracted more funding, which helps further refine emerging theories and improvements related to structural inference of dynamic causal effects. The impact on macroeconomic theory will undoubtedly be a positive one.

Keywords

Inflation, macroeconomics, ISIDCE, causal inference, econometrics

Discover other articles in the same domain of application