Barro–Gordon revisited: Reputational equilibria with inferential expectations

Abstract

We incorporate inferential expectations into the Barro-Gordon model (1983a) of time inconsistency and consider reputational equilibria. The range of sustainable equilibria shrinks as the private sector becomes more belief-conservative.

The Great Recession and the Two Dimensions of European Central Bank Credibility

“…In many circumstances it might be natural to assume that the IE test statistic of a central bank is some function of the departure of inflation from the target (here, the 2% ECB target). For example, this was the assumption of Henckel et al (2011). However, in the current crisis, it is both realistic and convenient to assume that the danger to credibility does not stem from the recent inflation history.…”

“…We now ask how the estimated coefficients in Table 3 might address the central questions of the paper. 28 It is clear that official target credibility has not been a serious issue in the GR. A number of countries have seen increases in the null hypothesis inflation rates (Austria, Belgium, Finland and Germany) but the remaining countries have seen decreases, some substantially.…”

“…The Cochrane-Orcutt representations for Portugal and Ireland did not converge, so we used the original (OLS) standard errors. 28 We will cast this discussion in terms of Table 3, though all the main points apply to the models of Table 2 as well, i.e. they are robust to our model simplification choices.…”

“…We eschew this explanation, which we find implausible, and instead propose that the increased variance of inflation forecasts could be explained by changes in beliefs about central bank credibility. 4 We model beliefs about central bank credibility by regarding the central bank inflation target as a null belief in a hypothesis test (Henckel et al, 2011), which can be modeled using the inferential expectations (IE) framework (Menzies and Zizzo, 2009).…”

“…Alternatively, an inflation bias may be present, implying a wedge between the official inflation target and the actual equilibrium value, and this is imperfectly observed by experts. This latter possibility is in the spirit of Gordon (1983a, 1983b) andHenckel et al (2011) 13. This has the considerable advantages of model parsimony and of nesting RE as a special case of IE, for when  = 1 the null must be rejected and RE obtains.…”