Detecting and testing causality in linear econometric models

Faliva, Mario; Zoia, Maria Grazia

doi:10.1007/bf02589041

Cited by 4 publications

(5 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is precisely the type of structure usually assumed in econometric models, except that in some models the above structure contemplates the existence of feedback reactions (see Faliva, 1992;Faliva & Zoia, 1994). A very clear and instructive correspondence between typical kinds of graphical models and (parameters of) linear models is commented by Cox & Wermuth (2004).…”

Section: Directed Acyclic Graphs and Structural Equation Modelsmentioning

confidence: 96%

Causality and Causal Models: A Conceptual Perspective*

Frosini¹

2006

Int Statistical Rev

View full text Add to dashboard Cite

This paper aims at displaying a synthetic view of the historical development and the current research concerning causal relationships, starting from the Aristotelian doctrine of causes, following with the main philosophical streams until the middle of the twentieth century, and commenting on the present intensive research work in the statistical domain. The philosophical survey dwells upon various concepts of cause, and some attempts towards picking out spurious causes. Concerning statistical modelling, factorial models and directed acyclic graphs are examined and compared. Special attention is devoted to randomization and pseudo-randomization (for observational studies) in view of avoiding the effect of possible confounders. An outline of the most common problems and pitfalls, encountered in modelling empirical data, closes the paper, with a warning to be very cautious in modelling and inferring conditional independence between variables.

show abstract

Section: Directed Acyclic Graphs and Structural Equation Modelsmentioning

confidence: 96%

Causality and Causal Models: A Conceptual Perspective*

Frosini¹

2006

Int Statistical Rev

View full text Add to dashboard Cite

show abstract

“…Only once the model has been duly estimated, can the coefficients of matrix C be properly evaluated. At this point, it proves useful to devise a procedure for testing the significance of the estimated loops (see Faliva and Zoia, 1994). To this end, let us observe that, once the matrix including all the feedbacks operating in the model…”

Section: Testing the Significance Of Feedback Loopsmentioning

confidence: 99%

“…a test for the significance of the loops can be based on the exam of the statistical non-nullity of the elements of matrix Γ * F which, unlike C, does not require the preliminary split of Γ into its components, given the feedback loops C + Ψ 1 and causal links Ψ 0 . In this context (following Faliva and Zoia, 1994), it can be proved that the j-th row of matrix Γ * F measures both the direct effect of the RHS endogenous variables on the j-th one and the feedback effect of the latter on the former variables. In fact, the direct effects of the RHS endogenous variables, collected in vector yo, on variable y j are included in the j-th row of matrix Γ (excluding its j-th element), that is…”

Section: Testing the Significance Of Feedback Loopsmentioning

confidence: 99%

“…Starting from this premise and following Strotz and Wold, 1960;Wold, 1964; and more recently Faliva, 1992;Faliva and Zoia, 1994); in this paper we have devised an operational method to distinguish the causal from the interdependent equations of a SEM. Other approaches for detecting feedback-loops arising in deterministic (error free) models are based on either graph or system theory (see e.g., Gilli 1992).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Identifying and Testing Recursive vs. Interdependent Links in Simultaneous Equation Models via the SIRE Package

Vacca¹,

Zoia²

2019

The R Journal

View full text Add to dashboard Cite

Simultaneous equation models (SEMs) are composed of relations which either represent unidirectional links, which entail a causal interpretation, or bidirectional links, due to feedback loops, which lead to the notion of interdependence. The issue is of prominent interest in several respects. Investigating the causal structure of a SEM, on the one hand, brings to light the theoretical assumptions behind the model and, on the other hand, pilots the choice of the befitting estimation method and of which policy to implement. This paper provides an operational method to distinguish causal relations from interdependent ones in SEMs, such as macro-econometric models, models in ecology, biology, demography, and so forth. It is shown that the causal structure of a system crucially rests on the feedback loops, which possibly affect the equations. These loops are associated to the non-null entries of the Hadamard product of matrices encoding the direct and indirect links among the SEM dependent variables. The effectiveness of feedbacks is verified with a Wald test based on the significance of the aforementioned non-null entries. An R package, SIRE (System of Interdependent/Recursive Equations), provides the operational completion of the methodological and analytic results of the paper. SIRE is applied to a macroeconomic model to illustrate how this type of analysis proves useful in clarifying the nature of the complex relations in SEMs.

show abstract

“…namely the partial derivatives of the functions in the right-hand side of the equation system, with respect to the current endogenous variables, and the covariance matrix of contemporaneous disturbances (for details, see [15][16][17]). As shown in the aforesaid references, the diagnostic about the causal or interdependent nature of a model is provided by the topological structure of the matrices above, or more conveniently by the corresponding binary matrices, written D b f and b , which can be assumed to be known a priori.…”

Section: Causal Structure Analysis: Backgroundmentioning

confidence: 99%

Causality and interdependence analysis in linear econometric models with an application to fertility

Barbieri

2013

Journal of Applied Statistics

View full text Add to dashboard Cite

This paper is an applied analysis of the causal structure of linear multi-equational econometric models. Its aim is to identify the kind of relationships linking the endogenous variables of the model, distinguishing between causal links and feedback loops. The investigation is first carried out within a deterministic framework and then moves on to show how the results may change inside a more realistic stochastic context. The causal analysis is then specifically applied to a linear simultaneous equation model explaining fertility rates. The analysis is carried out by means of a specific RATS programming code designed to show the specific nature of the relationships within the model.

show abstract

Detecting and testing causality in linear econometric models

Cited by 4 publications

References 11 publications

Causality and Causal Models: A Conceptual Perspective*

Causality and Causal Models: A Conceptual Perspective*

Identifying and Testing Recursive vs. Interdependent Links in Simultaneous Equation Models via the SIRE Package

Causality and interdependence analysis in linear econometric models with an application to fertility

Contact Info

Product

Resources

About