Search Results

This chapter explores a set of identifying restrictions on the short-run dynamics that might be used to supplement the long-run restrictions if the model is to be used to investigate the effect of economically-meaningful shocks. Particular attention is paid to modelling monetary policy decisions, inflation targeting, the derivation of a base rate reaction function, and the identification of monetary policy shocks.

This chapter describes the methodological confusions in econometrics via the growing labyrinth of the error term specification and interpretation. It traces the specification and interpretation from the pre-CC years and follows their evolution through the development of the three major reformative movements. It shows how various structural modelling strategies based on the CC tradition have tried to relegate the dynamics of time-series data to nuisance, atheoretical complications. In particular, it shows how the VAR approach gets somehow lost in of the dual interpretation of the error term as innovation error and structural shock. In contrast, the LSE approach seeks to decompose dynamic shocks into several components in an effort to keep the resulting error terms as atheoretical residuals free of statistical complications.

There are significant differences in unemployment levels between occupations, regions, age-groups, and races. These differences are very persistent and do not reflect the legacy of structural shocks. They are however closely connected to differences in turnover rate with differences in unemployment durations playing a small role. Unemployment level differences between age-groups are affected by demographic factors. But unemployment differences between occupations and regions can only be explained jointly in terms of mobility between groups.

This chapter examines the rise of the VAR (Vector AutoRegressive) approach. It shows that the VAR approach arises from a fusion of the CC tradition and time series statistical methods, catalysed by the rational expectations (RE) movement, that the approach offers a systematic solution to the issue of ‘model choice’ bypassed by CC researchers, hence essentially inheriting and enhancing the CC legacy rather than abandoning or opposing it. By tackling model choice, however, the VAR approach helps reform econometric research by shifting its focus from measurement of given individual structural parameters within theoretical models to identification/verification of more data-coherent theoretical models.

This chapter first introduces impulse response analysis and forecast error variance decomposition for unrestricted vector autoregressive (VAR) models and discusses the orthogonalized and generalized impulse response functions. It then considers the identification problem of short-run effects in a structural VAR model. It reviews Sims' approach and investigates the identification problem of a structural model when one or more of the structural shocks have permanent effects. Exercises are provided at the end of the chapter.