Click here to download

Behavioural models for manufacturing firms: analysing survey data
Author/s: Luciana Crosilla, Marco Malgarini 
Year:  2011 Issue: Language: English 
Pages:  25 Pg. 139-163 FullText PDF:  341 KB
DOI:  10.3280/POLI2011-004005
(DOI is like a bar code for intellectual property: to have more infomation:  clicca qui   and here 

Survey data on manufacturing firms are usually analysed on an aggregate basis, calculating for each question the so-called "balance" between the number of positive and negative replies. The simple average of selected balances is then commonly used to calculate the confidence indicator. While balance and confidence indicators provide an easy-to-compute and easy-to-understand quantification of survey results, and therefore are considered useful tools to analyse the sector’s cyclical situation, a cyclical analysis based on balance and confidence indicators alone fails to fully exploit all the information embedded in the survey. More specifically, computation of the balance statistic disregards "neutral" answers to survey questions and no attempt is made to identify potential relationships between the different responses to the various survey questions given by the same firms. A more in-depth study of this information can provide interesting insights into firms’ opinions on the economic situation. The contribution presents a new methodology based on cluster analysis that takes into account also the neutral answers and then uses it to assess the similarities and differences between the recent crisis and current recovery, and to compare these to past cyclical crises, specifically, the major recession of 1992-1993 and subsequent recovery in 1993-1995.
Keywords: Business cycles, business survey data, cluster analysis
Jel Code: C32, E32

Luciana Crosilla, Marco Malgarini, in "ECONOMIA E POLITICA INDUSTRIALE " 4/2011, pp. 139-163, DOI:10.3280/POLI2011-004005


FrancoAngeli is a member of Publishers International Linking Association a not for profit orgasnization wich runs the CrossRef service, enabing links to and from online scholarly content