Written by Michael McAlary
Service Area: Behavioural Analysis and Economics
Industry Sector: Government
Structural Path Equation Modelling
Chairmont has an enhanced diagnostic that overcomes the rudimentary weaknesses of traditional statistical methods. This will provide greater understanding of relationships and transmission mechanisms between the independent variables.
The widely used applications of Structural Equation Modelling (SEM) analysis typically focus on observing the simple 1 on 1 relationships between “independent” and “dependent” variables in order to understand cause and effect transmission dynamics. An example in the financial services industry might be the relationship between interest rate increases and home loan arrears. Chairmont has an enhanced diagnostic that overcomes the rudimentary weaknesses of such traditional methods. This modelling approach is referred to as Structural Path Equation Modelling (SPEM) which introduces a third variable type of a "mediating" variable that sits between dependent and independent variables.
Mediating variables may increase or decrease the impact of an independent variable on a dependent variable, as measured through multiple elasticity metrics. Under a SPEM framework the diagnostic applies behavioural measurement techniques in identifying these mediating variables through the use of psychological based statistical collection methods. In 2010 Chairmont conceptually applied its diagnostic to the Government’s 2008/2009 Stimulus Package which included cash transfers of $12.2bn. The diagnostic was aimed a better understanding the effectiveness of the cash transfers through the relationships and transmission mechanisms between the independent variables (who got what - amount by recipient type) and the dependent variables (impact on economy - GDP, Retail Sales, Unemployment). Chairmont addressed this through the identification of the mediating variables, i.e. the behavioural patterns of the stimulus recipients, and the associated economic flow-on effects from those behaviours to the dependent variables.
Click on the links above to view how this has been applied to the Federal Government's stimulus package.