### Discussion Paper

## Abstract

While there seems to be a well-established consensus about the underlying causes to the Greek crisis, less is known about internal and external transmission mechanisms that ultimately caused unemployment to increase rapidly over this period. Motivated by the structural slumps theory in Phelps (*Structural slumps*, 1994), the paper attempts, therefore, to uncover the dynamic mechanisms behind prices, interest rates, and external imbalances that contributed to the severity and the length of the crisis. The authors find that the strongly increasing real bond rate and unemployment rate together with a persistently appreciating real exchange rate and a deterioration of competitiveness in the euro-zone have contributed to persistently growing structural imbalances in the Greek economy. As the lack of confidence in the Greek economy grew steadily, the scene was set for a monumental structural slump. Over the crisis period, all variables exhibited self-reinforcing feedback adjustment somewhere in the system except for inflation rate. Unemployment took the burden of adjustment when the bond rate sky rocketed, competitiveness deteriorated, and confidence fell.

## Comments and Questions

Report for: “The Greek crisis: A story of self-reinforcing feedback mechanisms”

The study builds on Phelps (1994) to uncover the dynamic mechanisms behind prices, interest rates, and external imbalances of the Greek economy. The authors find that the increasing real bond rate and unemployment rate together with the real ...[more]

... exchange rate and a deterioration of competitiveness zone have contributed to persistently growing structural imbalances in the Greek economy.

Over the crisis period, all variables exhibited self-reinforcing feedback adjustment. Unemployment took most of the adjustment.

The authors employ Greek data and they build a cointegrated VAR framework. The paper is timely, of interest and well written.

Comments

1. This paper provides a very interesting overview of the literature on the Greek crisis. The authors could also extend by considering a few more studies:

Sarafidis, V. Panagiotelis A., Panagiotidis T. (2017) When did it go wrong? In The Greek debt crisis edited by C. Floros and I. Chatziantoniou.

2. Figure 1: is the inflation rate seasonally adjusted?

3. The authors might need to acknowledge that some of the variables of interest are operating at a different frequency (for instance bond rates (or spreads)).

Overall the paper is of high quality.

Thank you very much for your kind report and for the useful reference we had not been aware of! It certainly adds interesting information on how the Greek economy performed under this sample period compared to EU15 and the IIPS (Ireland, Italy, Portugal and Spain). As I am planning to ...[more]

... do an empirical/econometric comparison based on this group of countries, the reference will be highly useful for future work. But we shall with pleasure add the suggested paper to the reference list of our paper.

Inflation is not seasonally adjusted. We account for the strong seasonality by including seasonal dummies in the model. As appears from Figure 2 in the Appendix, there are still some autocorrelation left in the inflation rate residuals suggesting that the seasonal pattern might have been changing to some extent over the sample period.

One more comment: the bond rate is given as annual interest rates, but has been translated to monthly interest rates to be measured in a similar magnitude as the inflation rate.

This paper combines Phelps's (1994) structural slumps theory with the concept of imperfect knowledge expectations (Frydman and Goldberg, 2007) to explain the main determinants of the Greek crisis. The econometric analysis is based on a cointegrated VAR model with I(2) components applied on the period 2004:5-2017:1.

The explanation of the ...[more]

... Greek crisis is based on an interesting interpretation of the adjustment dynamics featured by the I(2) model, i.e. the interplay by error-correction and error-increasing mechanisms. which generates self-reinforcing feed-back patterns reflecting the deep and prolonged imbalances experienced by the economy.

One merit of the cointegrated I(2) model is that it gives rise to a very rich dynamic structure of adjustment, but the challenge it raises is how to give a meaningful structural interpretation to the somewhat "unconventional' dynamics it generates. This paper does an excellent job in this direction. The explanation of the Greek crisis is reasonable and tied to the policy implications. It clearly remarks which are the costs of prolonged periods of imbalances faced by small-open economies. The analysis seems to indissolubly rely upon the I(2) interpretation of the data, which is not totally obvious from the analysis in Section 3.2. The "wild fluctuations" of the Greek economy over the period 2004:5-2017:1 suggest that a VAR specification based on time-varying parameters could equally approximate the dynamics of the systems' variables. So, it would be interesting to complement the results and conclusions of the article with a robustness exercise showing that the main ingredients of the structural slumps theory interpretation advocated by the paper remain valid also outside the I(2) framework.

Thank you for your kind comments!

I agree that the data are extreme during the crisis and I admit to being positively surprised how well the I(2) model worked with only a few dummies needed to control for the most extreme shocks. When this is said, the I(2) model should ...[more]

... be considered a first order linear approximation to an underlying non-linear model. The problem with non-linear models in general is of course that the nonlinear features can be modeled in numerous ways and it is not always easy to know which is the best specification. Since the I(2) model worked reasonably well it suggests that the first order linear approximation can explain most of the variation in the data. Of course it would be exciting to see if one could do much better using a nonlinear approach. I would like to invite anyone (also you!) with experience of nonlinear modelling to use the paper's data for this purpose.

see attached file

See enclosed file!