Collinearity is a common problem in econometric models which can have significant implications on the stability and precision of structural analyses. In this paper, we explore the effects of collinearity on Vector Autorregressive (VAR) models by specifically testing the existence of the so-called pass-to-the-lag collinearity effect. This is of special relevance for structural econometrics as it involves modeling economic relationships and estimating parameters to understand the underlying structures. The results show that, indeed, correlation between two random variables can distort structural analyses through the aforementioned pass-to-the-lag effect, as it can transmit collinearity to the coefficients of their lagged values, causing problems in Impulse Response Functions (IRFs), Forecasted Error Variance Decompositions (FEVDs), Historical Decompositions (HDs), and statistical inference. Therefore, it is crucial to use appropriate techniques to detect and correct collinearity in this kind of models. By doing so, we can improve the accuracy of structural analyses and obtain more reliable and useful results for policy and decision making.