In 2020, the European Central Bank (ECB) started the Consumer Expectations Survey (CES) to gather information on a variety of crucial economic topics relating to household economic behaviour. Consumer inflation expectations have received particular attention because of their influence on realized inflation and the effectiveness of monetary policy. Unlike the European Commission consumer survey, which only publishes aggregated data, CES micro data are released on monthly basis, allowing for longitudinal data analysis. The data set consists of an ordinal variable recorded for a number of subjets at multiple time periods. The CUB model offers a flexible mixture distribution to represent ordinal data at a certain time point. In this paper, we modify that formulation to take time dependency of parameters into account. Since this class of models allows for both subjects' and objects' covariates, we exploit a CUB time varying model to examine consumers' attitudes on inflation across various respondent clusters.
A time varying CUB model for repeated measurements / Corduas, M., Di Iorio, F., Piccolo, D.. - 3:(2026), pp. 45-50. [10.1007/978-3-032-30881-8_8]
A time varying CUB model for repeated measurements
Corduas M.
;Di Iorio F.;Piccolo D.
2026
Abstract
In 2020, the European Central Bank (ECB) started the Consumer Expectations Survey (CES) to gather information on a variety of crucial economic topics relating to household economic behaviour. Consumer inflation expectations have received particular attention because of their influence on realized inflation and the effectiveness of monetary policy. Unlike the European Commission consumer survey, which only publishes aggregated data, CES micro data are released on monthly basis, allowing for longitudinal data analysis. The data set consists of an ordinal variable recorded for a number of subjets at multiple time periods. The CUB model offers a flexible mixture distribution to represent ordinal data at a certain time point. In this paper, we modify that formulation to take time dependency of parameters into account. Since this class of models allows for both subjects' and objects' covariates, we exploit a CUB time varying model to examine consumers' attitudes on inflation across various respondent clusters.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
