“Twin deficits” in Greece: in search of causality

Abstract:

The paper discusses the trajectories of the Greek public deficit and sovereign debt between 1980 and 2010 and its connection to the political and economic environment of the same period. We pay special attention to the causality between the public and the external deficits in the period after 1995, the post–Maastricht Treaty period. We argue that, due to the European monetary unification process and the adoption of the common currency, causality ran from the external deficit to the public deficit. This hypothesis is tested econometrically using both Granger causality and cointegration analyses. We find empirical support for this hypothesis.

Key words::

• crisis
• Greece
• imbalances
• public debt
• twin deficits

JEL classifications::

• E62
• F21
• F34
• F41

Notes

¹The terminology of closures has been popularized by Taylor and Lysy (1979) to denote the set of variables that are considered endogenous and exogenous and the different causalities that can be placed in a set of restrictions or simple economic relations.

²The seminal study in support of the twin-deficits hypothesis for the United States in the 1980s is by Abell (1990), who shows by means of a VAR model that increases in the fiscal deficit led to an appreciation of the currency and thus to a deterioration of the trade deficit. Recent studies have also found empirical evidence against the twin-deficits hypothesis. Erceg et al. (2005) find that the fiscal deficit has a relatively small effect on the trade balance in the United States. Baharumshah et al. (2006) find a two-way causality relationship for the vast majority of countries in their sample.

³For a review of gap models, see Taylor (1994).

⁴The first to use this equation to analyze the dynamics of government debt was Evsey Domar (1944).

⁵The economic rationale behind this focus on low inflation is that in a world with rational expectations, unemployment will always be at its natural level and any effort to lower unemployment below that level will only create inflation without any employment gains. This is the famous argument of rules rather than discretion made by Kydland and Prescott (1977) and Barro and Gordon (1983). The rest of the Maastricht criteria and the institutions of the EMU were built around this target of low inflation. For example, the second criterion of low government deficit (below 3 percent) or the call for a conservative central banker are the textbook policy conclusions of the aforementioned economic models because they create an institutional structure that prohibits discretionary policy.

⁶The hard-drachma policy rationale is explained in a paper written by the deputy director and the research director of the Bank of Greece six years later, after the successful entry into the eurozone (Garganas and Tavlas, 2001). Their analysis is based on the 1997 World Economic Outlook of the IMF titled “Exchange Rate Arrangements and Economic Performance in Developing Countries (IMF, 1997).

⁷In that sense this process can be viewed (to paraphrase Keynes) as one bubble in the whirlpool of bubbles of this period. It is characteristic that the latest edition of the seminal Manias, Panics and Crashes identifies this process in Greece as part of one of “The big ten financial bubbles” (Kindleberger and Aliber, 2011, p. 11).

⁸The exchange rate refers to the left vertical axis while the external finance to the right one.

⁹An exception to this pattern in the first period is 1981. This is probably due to the recording of the capital transfers related to the entry of Greece to the European Economic Community the year before. Data from other sources, such as the World Economic Outlook of the International Monetary Fund (2012) show an increase in the current account deficit for that year.

¹⁰The volume is the product of collaboration between the Bank of Greece and the Brookings Institution.

¹¹Since the power of the Dickey–Fuller test is weak, especially in small samples, we pursue a cointegration analysis below for both periods.

¹²Note that we only present the estimation results of the two-dimensional system comprising net exports and the public deficit, since adding variables such as the real output or the inflation rate as in Vamvoukas (1999) does not change the results substantially.

¹³The only exception is the trace test for the specification with time trend for the period from 1980Q1 to 1994Q4, which indicates no cointegration relation.

¹⁴For the pre-1995 period our data do not contradict the results obtained by Vamvoukas (1999), who also tries to identify the direction of causality between the public and external deficits of the Greek economy using annual data from 1948 to 1994. He employs a trivariate Granger causality analysis as well as a trivariate cointegrated vector autoregression model including the public deficit, the external deficit, and real output or inflation. According to his findings, the causality is running from the public to the external deficit. This is consistent with the view that pre-1995 external deficits may have been driven by public deficits.

¹⁵Some specific policy solutions for Greece are provided by Papadimitriou et al. (2013, 2014).

Additional information

Notes on contributors

Michalis Nikiforos

Michalis Nikiforos is Research Scholar at the Levy Economics Institute of Bard College.

Laura Carvalho

Laura Carvalho is Assistant Professor at São Paulo University, São Paulo, Brazil.

Christian Schoder

Christian Schoder is Assistant Professor at Vienna University of Economics and Business, Vienna, Austria.