Greece and the Troika--Lessons from International Best Practice Cases of Successful Price (and Wage) Adjustment.

Author:Belke, Ansgar
  1. Methodology

    In this paper we present three international best practice (or those considered as benchmarks) cases of successful price and wage adjustment. In this context we critically evaluate the reforms undertaken within the Greek Adjustment Program (see, for instance, Alcidi et al., 2014).

    Among the benchmark practice cases there will be Australia with its flexible exchange rate (Wood, 2013), Latvia with its exchange rate peg (Alcidi and Gros, 2013, Gros et al., 2014, and Sippola, 2011), and the East German "Laender" after reunification and entering into a currency union with Western Germany (Wolf, 2011).

    The labour market is a key place where the adjustment takes place in a context of fixed exchange rates, irrespective of whether in the form of a currency board or a monetary union. Given that under this assumption external (currency) devaluation cannot act as safety valve to regain competitiveness and reduce external imbalances, the economy is forced into internal devaluation. This implies that prices must adjust and wages are a key price. Wage adjustment is key to allow the economy reach both internal and external balance. The immediate reaction after a sudden stop to capital inflows is of course that domestic demand falls. In the short run this is the only way to eliminate a current account deficit since it takes time to increase exports. A fall in wages will actually accelerate the fall in GDP in the short run. But the fall in wages stimulates net exports, allowing the economy to grow without incurring again large current account deficits (Belke and Gros, 2017). Without a fall in wages, domestic demand has to remain depressed and unemployment would remain higher. We therefore propose to analyse cases of significant wage/price adjustments, which could serve as a benchmark for Greece. (1)

    Australia provides an extreme example, with a formal wages and price policy, accompanied by strong competition policies and legal authority to police the adjustment and to impose substantial fines where the basic rules are not followed. This package was adopted by the Australian government in 1982 to deal with excessive real wage cost levels, inflation and high unemployment under flexible exchange rates. Some argue that the policy was overall successful: real wage costs subsequently fell by 12 per cent, inflation fell back and full employment was restored (see, for instance, Wood, 2013).

    Such a policy of extensive controls seems no longer feasible today. European countries on fixed exchange rates thus had to rely on a so-called austerity-led internal devaluation strategy. This was the case for Greece, Portugal, Latvia and other program countries. We thus propose to look at Portugal and Latvia as real life comparators. Another useful benchmark is the case of Eastern Germany, because it shows how the adjustment could take place within one country, which one can consider as the extreme case of a monetary union.

    Latvia, which maintained a tight peg to the Euro and thus, was also forced to adjust without any devaluation, provides a particularly important comparison because the initial optimism that the Greek program should succeed was based on the perceived success in this case. The macroeconomic adjustment program for Latvia was much stricter than that for Greece, but the adjustment was quicker and followed by a stronger rebound. At the trough of Latvia's recession, the program was also off-track and failure seemed imminent, but it turned out that the sharper-than-planned adjustment cleared the way for a solid recovery (Alcidi and Gros, 2013, Biggs and Mayer, 2014, Gros et al., 2014). (2)

    Portugal, which came shortly after Greece, provides another benchmark as the country seemed to face similar problems as Greece in terms of low competitiveness (for details see Alcidi et al., 2014).

    As another best practice case we refer to the new states following German reunification. Whereas this case is informative in suggesting that significant internal devaluations in larger, relatively closed economies are feasible, it of course does not speak in favour of the desirability or effectiveness of internal devaluations in the Eurozone crisis economies. Such an assessment would require close attention to the specific circumstances, notably the cross-linkages between the external cost competitiveness and the fiscal challenge, as well as to the likelihood that a cost competitiveness gain would result in a significant export response--which does not seem to be the case for Greece a priori (Paque, 2009, and Wolf, 2011).

    There exists a substantial literature (see Belke and Gros, 2017) on what constitutes a 'success' of IMF programs. (3) The current policy debate puts emphasis on the trade-offs between "external" and "internal" devaluation. Obstfeld (1997) claims that the former can be successful even in the presence of real wage resistance. In this context Belke and Gros (2017) and Alcidi et al. (2014) find it interesting to compare what happens to real wages under "external" and "internal" devaluation. The EU provides many examples of both adjustment methods. But our concern is not to judge whether IMF programs are well designed. We take a simpler, policy orientated view: we regard a program as a 'success' if the country concerned gains market access within the time frame considered of the program itself (and if there is no need for another program within a short period of time). The case of Greece stands out in this respect (Belke, 2017). But market access should of course not constitute the only criterion of success. But market access usually comes only if the country grows again. Greece was not able to exit its program because the country did not manage to embark on a sustainable growth path (Belke and Gros, 2017).

    The next section looks at wage developments at the macroeconomic level.

  2. Wage developments at the macroeconomic level

    The relationship between slack in the labour market and wage growth is usually called the Phillips curve. (4) The basic idea is simply that an excess supply of labour (measured by the unemployment rate) should lead to lower wage growth.

    This mechanism seems to have operated in Greece as nominal wages have fallen while unemployment rose above 20 % as shown in Figure 1 below. However, a closer look at the data shows that prior to 2009 there was no clear relationship between unemployment and wage growth in Greece. The data up to 2008 (in the red oval) shows no clear relationship between these two variables.

    This absence of a Philips curve type relationship pre-crisis sets Greece apart from other program countries with a low per capita income level.

    Figure 2 below shows that in Portugal wages had already begun to decelerate as unemployment increased well before the crisis broke. The increase in unemployment and the further fall wage growth during the acute phase of the crisis seems to have represented a continuation of a cyclical movement, which had already started beforehand.

    The same can be said of Latvia, where wages had been growing at an extreme pace (an increase of close to 40 % at the peak of the boom in 2007), but decelerated sharply during the bust, which wage growth falling to minus 10 % in 2009 (while unemployment increased by 12 percentage points (from about 6 to 18 %).

  3. International best practice cases of successful price and wage adjustment--a review

    After having discussed briefly the three program countries with fixed exchange rates we now turn to a more in depth discussion of two more complex cases--one under flexible exchange rates with an inflation problem (Australia) and another one within a monetary union with unified labour market institutions (the German New "Laender").

    3.1. Australia

    The "Prices and Incomes Accord"

    In 1982, the Australian government adopted an incomes policy (5), consisting of formal wages and prices policies, called the "Prices and Incomes Accord" to deal with excessive real wage cost levels, inflation and high unemployment (apparently also core problems in Greece at the start of the crisis), allegedly without reducing the living standards of Australians --under flexible nominal exchange rates. Notably, this represents a different restriction than that faced by Greece today, i.e. "irrevocably" fixed nominal exchange rates within a currency union. The Accord represented an agreement met in 1983 between trade unions and the Australian Labour Party government (Prime Minister Bob Hawke and Treasurer and later Prime Minister Paul Keating) which, however, did not include the employers. All participants were obliged to relate their decisions to the overall economic situation and outlook (Romanis Braun, 1986, p. 3f.). Unions agreed to restrict wage demands, and the government in turn pledged action to minimise price rises (potentially useful also in the Greek case).

    The rationale for resorting to this measure was that the relationship between wage inflation and unemployment had deteriorated considerably, as shown in the chart below. For example, by 1974, wages were increasing by close to 30 % p.a. although unemployment stayed at 2 %, about the same level as the mid-1960s, when wages were increasing only by about 5 %. Wage inflation then decelerated considerably to about 5 % by 1983, but at that time unemployment had increased to 10 %. The trade-off had thus considerably worsened. The Accord can be seen as an attempt to shift the curve back towards the origin (i.e. to reach lower (wage) inflation with lower unemployment).

    The Accord, which was renegotiated several times, can be characterised as a strategy of a formal wages and prices policy composed of temporary, tailor-made wages and prices policies. Among its formal...

To continue reading