Unit labor costs and the Greek current account balance

(1)

1

Bachelor Thesis

University of Amsterdam

Faculty of Economics and Business Study: Economics

Track: Economics Author: Jordi Fierlings

Email: Jordifierlings@hotmail.com Student number: 10540199 Supervisor: Egle Jakucionyte Second Reader: .

Date: January 30, 2017

Unit labor costs and the Greek current

account balance

(2)

2

1. Introduction

The sovereign debt crisis in Greece that originated in the first half of 2010 was highly covered in the media worldwide. A deterioration of the competitive position of Greece which up till then had been covered by a satisfying economic performance came to light. The deterioration of their competitive position can mainly be attributed to a large increase in unit labor costs compared to other members of the Euro zone (De Grauwe, 2012). As a consequence, Greece accumulated a large current account deficit. This led to a significant increase in government and external debt and this in turn was one of the major causes of the European sovereign debt crisis (Pisano-Ferry, 2012a). As of early 2010, Greece was unable to raise new funds on the bond market which did not allow for the government to satisfy its debt obligations. In order to prevent the crisis from worsening, it was granted a number of support loans. In exchange for the financial support, the European countries and the IMF imposed an adjustment program on Greece in may 2010. This program required the country to restructure its economy and was aimed at improving competitiveness.

In the policy debate and literature that followed, it was decided that internal devaluation should be the main pillar of the program. In other words, Greece should aim at restoring price competitiveness by means of productivity improvements and/or wage cuts (Darvas 2012). This price competitiveness channel consists of the process of wage setting, price setting, as well as the impact of price competitiveness on exports and on the overall spending in the economy. The ultimate aim is to lower unit labor costs. Through the competitiveness channel, the export sector is expected to become the driving force of economic growth. This should then lead to an improvement of the current account balance. An economy can thus improve its price competitiveness via internal devaluation by reducing unit labor costs with positive spill over effects on its exports which will result in a positive effect on the current account balance (Alexiou, 2013).

The hoped for success of such a policy is based on the assumption that changes in the overall price level are brought up about by changes in unit labor costs. A decrease in unit labor costs would thus have a negative impact on inflation, making Greek products more competitive compared to the economies which did not lower their unit labor costs. Indeed, European countries which saw their unit labor cost grow relatively strong have also experienced higher inflation than countries which had a more modest unit labor cost development. On the other hand, countries in which the increase in unit labor costs was relatively weak, saw a modest increase in prices.

(4)

4 Fig.1 : Inflation and Growth of Unit Labor costs in Euro Zone countries (1999-2008)

This paper will address the issue whether or not changes in unit labor costs affected the current account balance in Greece. The answer to this question is vital, considering the emphasis that the European commission has put on decreasing unit labor costs as a possible solution for improving the current account balance in Greece (EC, 2010).

The contribution of this paper consists of two aspects. First, this paper analyzes the effect that unit labor costs had on the current account balance. It does this for a specific country, Greece. The findings of this paper suggest that unit labor costs has a strong negative significant effect on the current account balance in Greece. Second, this paper makes a distinction between the period after the Greek structural reforms took place in 2010 and the period before the adjustment program was enacted. It then estimates the differences in the impact of unit labor costs between the two periods. It is concluded that the effectiveness of unit labor costs on the current account balance is higher from 2010 on, compared to the period before 2010. There can be various reasons for this, which will be discussed in more detail in chapter 3. The significant difference in impact of unit labor costs on the current account balance between the two periods has important policy implications. It implies that policy makers ought to change their methods and actions regarding unit labor costs and the current account balance when the economy is hit by a significant shock.

This paper will perform a OLS regression to analyze the relationship between unit labor costs and the current account balance in Greece. It will make use of time series date that run from the second quarter of 1996 to the second quarter of 2016. Also it will be tested whether changes in unit

(5)

5 labor costs affect the current account balances in Greece differently after the structural reforms of the adjustment program took place.

The next chapter will discuss the concept of unit labor costs and show its trend in Greece for a selected time period. Thereafter, it will define the current account balance and investigate the current account balance in Greece. Chapter 3 will focus on the relationship between unit labor costs and the current account balance. It will give arguments both for and against a causal relationship between the former and the latter. Subsequently, chapter 4 will empirically test the theory by making use of econometric analysis. At last, the fifth chapter will state the conclusion and provide a summary of the finding of this paper.

2. Definitions and developments of unit labor costs and the current

account balance in Greece

To be able to look at the causal relationship between unit labor costs and the current account, the separate terms have to be discussed individually. This chapter begins with explaining the concept unit labor costs and will then show the development of unit labor costs in Greece. This chapter will thereafter investigate the current account balance and provide arguments for its importance. At last, the development of the current account balance in Greece will be discussed.

2.1 Defining unit labor costs

Nominal unit labor cost are the ratio between wage costs per worker, or labor compensation, and the average level of productivity (ECB, 2017). They thus represent a direct link between productivity and the cost of labor used in generating output. A decrease in unit labor costs means a decrease in marginal costs which results in an increase in competitiveness. It is evident that firms want to produce at minimum costs and that they hence want to decrease their marginal costs. These costs can thus be decreased by either decreasing the cost of labor or by increasing productivity.

𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝑈𝑛𝑖𝑡 𝐿𝑎𝑏𝑜𝑟 𝐶𝑜𝑠𝑡𝑠 =_{𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑙𝑎𝑏𝑜𝑟 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑣𝑖𝑡𝑦}𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑙𝑎𝑏𝑜𝑟 𝑐𝑜𝑠𝑡𝑠 (Eq. 1)

The average costs of labor are determined by the nominal wage that an employer pays his workers and by other labor related costs. These other labor related costs include the contribution of employers to corporate pension schemes, severance pay, and the social security payments. This suggests that the government, via its labor market policies, plays an influential role in the development of the wage costs.

(6)

6 The average labor productivity is calculated by dividing the total output of the firm by the amount of workers. This number tells how many products have been made per worker on average. The unit labor costs thus show how many labor cost is required to produce one product.

When one wants to look at the unit labor costs for the whole economy, the equation becomes slightly different:

𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝑈𝑛𝑖𝑡 𝐿𝑎𝑏𝑜𝑟 𝐶𝑜𝑠𝑡𝑠 = _𝑎𝑙𝑝𝑎𝑙𝑐= 𝑎𝑙𝑐∗𝐿_{𝑎𝑙𝑝∗𝐿} = 𝑡𝑜𝑡𝑎𝑙 𝑙𝑎𝑏𝑜𝑟 𝑐𝑜𝑠𝑡𝑠_{𝑟𝑒𝑎𝑙 𝐺𝐷𝑃} (Eq. 2)

Alc is the average labor costs and alp is the average labor productivity. L is the number of workers in the whole economy. The ratio between the total labor costs and real GDP will give the unit labor costs in a nation.

2.2 Developments of unit labor costs in Greece

Figure 2 shows the growth in unit labor costs for the selected euro zone countries and for the euro zone as a whole from 1991 to 2010. The figure shows that unit labor costs have risen significantly in Greece, both in nominal terms as well as in relative terms vis-a-vis the individual euro zone

countries. This resulted in a loss of competitiveness, making Greek products more expensive on the international market. Note that during this period unit labor costs in Germany actually declined as a result of reforms in the German labor market undertaken at the beginning of 2002.

The unit labor cost growth in Greece may be the outcome of various factors and processes that are hard to disconnect from each other and difficult to measure. The difference in relative

70.00 80.00 90.00 100.00 110.00 120.00 130.00 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Fig. 2 : ULC developmens of selected Euro area economies 1999-2010

Germany Ireland Greece Netherlands Portugal EA19

(7)

7

productivity growth between the tradable and non-tradable sectors may have been a cause for this growth (Papademos, 2007). These differences in productivity growth may have caused wages to rise in sectors that did not experience a productivity growth. In Greece, productivity growth was highest in the tradable sector. This resulted in rising wages in the non-tradable sector which exceeded productivity growth in that sector, putting upward pressure on prices. This effect is called the Balassa-Samuelson effect (Papademos, 2007). Second, the European commission attributed the large growth in unit labor costs in part to spill over effects from increasing wages in the public sector (EC,2010). Third, the characteristics of the Greek economy were also a cause for strong unit labor cost growth. The Greek economy, in contrast to other countries in the Euro area, is dominated by a relatively large number of small- and medium-sized family firms (Oltheten, 2003). Oltheten argues that this prevented these firms from exploiting economies of scale. As a consequence, most Greek firms were not able to compete with foreign competitors and were thus mainly selling its products on the home market. As a consequence, there was no incentive for these firms to lower their costs of labor to make them more competitive. Fourth, the gradual expansion of the state and its social income policy which resulted in a sharp increase in labor costs. Fifth, the flawed structure of the Euro zone and the neglecting of the no bail-out clause1 by financial markets (Stiglitz, 2016). Stiglitz argues that the convergence of the interest rates between the core and the periphery countries, which resulted in a relatively large increase in wages in the periphery countries, is evidence of this neglecting process. The lower interest rates in the periphery countries namely boosted consumption which in turn increased the demand for labor and thus created upward-pressure on wages (Stiglitz, 2016).The growth in unit labor costs in Greece between 1999 and 2010 can thus be attributed to both structural and country-specific factors.

Figure 3 separates unit labor costs in its two components, compensation per employee and the labor productivity. The compensation per employee is being used instead of labor costs because of data limits. The figure shows that despite a very strong productivity growth, increases in

compensation per employee have resulted in a relatively high unit labor cost growth compared to the Euro area average. This was also the point that the vice president of the ECB Papademos (2007) made during his speech on ''inflation and competitiveness divergences in the euro area countries' for the ECB in 2007.

(8)

8 Figure 4 shows the period the structural reforms took place in 2010 in Greece. What is clear is that the countries that saw their unit labor costs rise between 1999 and 2010 are now on the other side of the spectrum. Unit labor costs in Greece declined by 25%, reflecting a sharp decrease in wage costs and a modest decrease in productivity.

The drop in wages in Greece can be attributed to both the high increase in unemployment and to the reforms that were part of the adjustment program that was imposed on Greece in 2010. The reforms discussed in this program were mainly focused on the public sector, where according to the European commission wages were too high in comparison to the wages in the private sector (EC, 2010). Cuts in labor costs in the public sector consisted of the abolition of bonuses for holiday periods. Also the allowances paid to high-wage earners were reduced and pension payments were

0 50 100 150 200

Fig.3 : Nominal unit labor costs and their components, Greece, 1999-2015

Compensation LProductivity NULC

Source: Calculation of author based on data from Eurostat

75 80 85 90 95 100 105 110 2010 2011 2012 2013 2014 2015 A xi s Ti tle

Fig. 4 : ULC developments of selected Euro area economies 2010-2015

Germany Ireland Greece Netherlands Portugal EA19

(9)

9 cut. The reforms of the second adjustment program which was imposed in 2012 now also focused on the private sector, because the reforms of the first program did not lower unit labor costs by the desired amount (EC, 2012). Several measures were adopted in relation to collective bargaining, so as to reduce the downward rigidity on wages. Also measures that promoted the level of competition were adopted. The government furthermore reduced the minimum wage by 22 percent, or even by 32 percent for those younger than 25. In addition, social contributions were reduced by 5 percent. Wages also dropped because of a rise in unemployment, which caused the supply of labor to increase.

The reforms of the two adjustment programs resulted in a modest decrease of productivity. Alexiou (2013) argues that this decrease can be attributed to a reduction in productive capacity and net capital formation in Greece. This hypothesis relies on the research done by Timmer and van Ark (2000), who found a significant positive relation between capital formation and productivity in South East Asia.

The decrease in productivity was thus offset by a larger decrease in wages, resulting in a lowering of unit labor costs.

2.3 Defining the current account balance

𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝐴𝑐𝑐𝑜𝑢𝑛𝑡 = (𝑆 − 𝐼) + (𝑇 − 𝐺) (Eq. 3)

Here S and I represent private saving and private investment respectively, which results in net private savings. T and G represent tax income and government spending, which is the fiscal surplus or net public savings. A country’s current account is thus the difference between its savings and its domestic investments or, likewise, between its exports of goods and services and its imports. If the sum of net public and private savings is positive we find a current account surplus and vice versa.

If an economy is running a current account deficit, there is either more domestic consumption or investment than that economy is producing. This state of the current account balance can only be sustained if some other economies are lending their savings to the domestic economy or when the domestic economy is running down its foreign assets. These economies that are lending their savings do this mostly in the form of debt. This means that countries with high current account deficits are accumulating debt.

There are two ways to investigate the current account and its changes over time, through the national savings identity and through international trade competitiveness (Krugman and

Obstefeld, 2015). Explaining changes in the current account via the national savings identity focuses on the flow of capital.

(10)

10 competitiveness. This focuses on the trade aspect of the current account and is called the

competitiveness approach. A country is expected to improve its current account when it becomes more competitive. The real exchange rate is generally used to measure cost and price

competitiveness (Lipschitz and Mcdonald, 1992).The real exchange rate measures the value of currencies, taking into account changes in the price level. The real exchange rate can thus change via adjustments in relative prices. The change in relative prices are brought up about by, among other variables, a change in unit labor costs (Kendrick, 1961). In the case of a monetary union with a common currency, only the change in relative prices influences the real exchange rate. Chapter 3 will discuss the theoretical relationship between competitiveness and the current account in more detail.

2.4 Why study the current account balance?

The question whether a large divergence between national imports and exports is a potential reason for concern dates back to the origins of economic theory. English writers who were concerned about the outflow of gold coins due to large imports introduced the expression ‘’balancing of trade’’ in the beginning of the 17th century(Price, 1905). The concept was further developed by Hume (1752) who argued that a large constant trade deficit and a large surplus were neither desirable nor feasible. Much of the debate since focuses on Hume's argument by stressing that the current account balance can be a potential channel for an international shock transmission, or a carrier of financial

vulnerability. According to Obstfeld (2012) it is therefore important to analyse large current account imbalances because it can signal macroeconomic and financial risks.

Blanchard and Milesi-Ferretti (2011) state that there are two main reasons why the current account needs to be investigated. First of all, the level of the current account can be seen as a predictor of future economic tumult. When the current account of a country becomes too high and unsustainable, international markets can lose their confidence in the government's ability to repay its debts. This in turn can result in considerable downgrades of the countries credit ratings and an increase of the interest rates of its loans. A country can lose access to the debt market if these interests rates become too high, paving the way for a crisis.There are historically various examples of crises that have been predated by a large current account deficit. The United States in 2007, Iceland in 2008, and Greece in 2010 are among the most recent examples. However, a mere incidental historical relationship does not assume causality. Catão and Milesi-Ferretti (2011) used data that covers 1970-2011 to research both advanced and developing economies and find that bigger current account deficits raise the likelihood of an external crisis. Furthermore, Pisani-Ferry (2012a) claimed that the current account imbalances were at the heart of the European sovereign debt crisis.

(11)

11 Gourinchas and Obstfeld (2012), Mendoza and Terrones (2008) and Schularick, and Taylor (2012) state in line with the previous that a rapid increase in domestic credit growth, which has to be matched by a deficit on the current account, is a statistically significant crisis predictor. In contrast, Reinhart and Reinhart (2009) state that there is mixed evidence that a large current account deficit in advanced economies raise the chance of a crisis. The literature is not consistent and therefore the subject needs to be researched further.

Secondly, significant changes in current account imbalances can have strong macro implications, both at home and abroad. The current account plays a vital role as a component of a country’s aggregate demand. When a country spends more than it earns in the form of income, this can only be sustained either by sales of foreign assets or by debt obtained from abroad. If foreigners for some reason stop providing the necessary liquidity, the current account must adjust by a large drop in aggregate demand. This scenario is what Calvo and Reinhart (2000) called the ''sudden stop''. It can involve adjustments in resources allocation and relative prices. The negative macro effects of a large current account can be especially harsh within a monetary union. A country which wants to decrease it current account deficit needs to increase its exports. This increase in exports can only be brought about by a depreciation of the real exchange rate. However, since the nominal exchange rate cannot change, the real exchange rate must change via a change in relative prices. Problems associated with a country's current account imbalance can thus hurt the home economy, and countries related to the home economy. Milesi-Ferretti (2011) namely argues that negative consequences of sudden changes in the current account also have spill-over effects to other economies. A crisis in one country can spread to other countries. This is exactly what happened in Europe during the European debt crisis (De Grauwe, 2012).

A large drop in aggregate demand can also affect the current account balance via the import channel (Krugman and Obstefeld, 2015). The drop in demand will namely lead to a drop in imports which will lead to an improvement of the current account balance. Paragraph 3.2 will discuss the relationship between aggregate demand and the current account balance in more detail.

Now that it has become clear that the current account balance is an important economic indicator, this chapter now continues with the examination of the current account balance in Greece.

(12)

12

2.5 Current account developments in Greece

Greece has historically been running significant current account imbalances. The persistence and scale however of these imbalances in the period following the introduction of the Euro in 1999 appears to have been greater than in earlier decades (Barnes, Lawson and Radziwil 2010).

Figure 5 shows the development of the Greece current account balance from 2002 to 2010. What is clear is that the current account deficits of Greece have been large and persistently above the Euro area average. De Grauwe (2012) argues that these current account deficits were too high to sustain and thus eventually lead to the Greek debt crisis, which resulted in the European debt crisis. The deterioration of the current account position of Greece can be attributed to several factors. Bitzis, Paleologos and Papazoglou (2008) empirically documented the most important drivers of the decline in the Greek current account balance. First, a gradual loss of competitiveness as a result of higher domestic inflation and relative unit labor costs. Second, a sharp rise in public deficits and debts after Greece joined the single currency area. The increase in debt was caused by superfluous demand by the government. This in turn increased the demand for imports and thus affected the current account balance negatively. Third, higher GDP growth, mainly driven by domestic demand, in the initial years following the adoption of the common currency and developments in world oil and freight prices, which may have also affected the country’s current account position during the period. In addition, Monokroussos and Thomakos (2012) add that the domestic financial deepening after the euro adoption also attributed to the deterioration of the current account balance.Financial liberalization and financial deepening are namely often associated with lower private saving, this is due to the relaxation of the intertemporal budget constraint facing households. Ceteris paribus, lower private savings lower the current account balance (see Eq. 3).

-20 -15 -10 -5 0 5

Fig. 5 : Current account deficit Greece, 2002-2010

Greece Euro area

(13)

13 Figure 6 shows the period after the adjustment program of 2010 was enacted. From the above we can conclude that the deficit substantially decreased during this period.

Theoretically, there are many variables that influence the current account. For a complete overview of the theoretical and empirical significant determinants of the current account balance, see (Chinn and Prasad, 2000; Gruber and Kamin, 2005; CPB, 2015). This paper takes into account that the current account is affected by many variables and takes one variable out of that list, namely unit labor costs, and researches the impact of this variable on the current account. The next chapter therefore addresses the issue of the theoretical relationship between unit labor costs and the current account balance.

3. Assessing the relationship between unit labor costs and the

current account balance

In the previous chapter, the concepts of unit labor costs and the current account balance have been discussed separately. Also the trends of these two concepts in Greece has been discussed. This chapter will now look at the theoretical relationship between unit labor costs and the current account balance. The goal of this chapter is to examine the influence that unit labor costs has on the current account. In order to test this hypothesis, this chapter will look at arguments that assume a causal relationship between the former and the latter. It will thereafter discuss arguments that do not assume this causal relationship. After reading this chapter it will be clear that a strong relation exists between the two concepts.

-20 -15 -10 -5 0 5 10 20 10-Q1 20 10-Q2 20 10-Q3 20 10 -Q4 20 11-Q1 20 11-Q2 20 11-Q3 20 11-Q4 20 12-Q1 20 12-Q2 20 12-Q3 20 12-Q4 20 13-Q1 20 13-Q2 20 13-Q3 20 13-Q4 20 14-Q1 20 14-Q2 20 14 -Q3 20 14-Q4 20 15-Q1 20 15-Q2 20 15-Q3 20 15-Q4 20 16-Q1 20 16-Q2

Fig. 6 : Current account deficit Greece, 2010-2016

Greece Euro area

(14)

14

3.1 Arguments supporting a causal relationship from unit labor costs to the current

account balance.

There are two different explanations on how unit labor costs might affect the current account balance. The first explanation regards overvalued prices or low competitiveness as a cause for decreasing current account balances and is called the competitiveness approach(Kennedy and Slok, 2005:9;EC, 2010; Krugman 2015;Shambaugh, 2012). This argument dates back to the Ricardian model (Ricardo, 1817). This model predicts that a country will produce the good in which it has a comparative advantage. Countries have a comparative advantage over others in producing a

particular good if they can produce that good at a lower relative marginal cost. These marginal costs are affected by the relative wage and productivity (see section 2.2). If the relative wage or

productivity would change, the amount of products that a certain country exports thus will also be affected. If the wage level in country A decreases and in country B stays constant, the model predicts that country A will increase its exports and that country B thus increases its imports. The increased exports of country A will have a positive effect on the current account balance of country A. A decrease in unit labor costs thus yield an improvement of the current account balance.

Golub and Hsieh (2000) find fairly strong support that Ricardian model performs well empirically. In the vast majority of cases that they researched, relative productivity and unit labor costs help to explain US bilateral trade patterns. Countries that have relatively lower unit labor costs are thus likely to have a positive current account balance.

Kennedy and Slok (2005) argue that a fall in a countries relative wage and prices is likely to affect trade flows rather quickly. Also the European Commission (2010) recognized the correlation between wages and prices on the one hand and the current account on the other. They argue that the increased unit labor costs in Greece eroded external competitiveness and that this resulted into a rapid growth of the current account deficit. In addition, Shambaugh (2012)states that a lack of competitiveness in a monetary union will eventually result in an overvaluation of its prices. He argues that lowering unit labor costs are an important way of reducing this overvaluation, which will in the long term improve the current account balance.

The relationship between the unit labor costs and the current account balance have also been empirically researched. Zemanek et al. (2010) analyze the role that private restructuring and structural reforms have on the current account balance by making use of a panel regression. They find that structural reforms and wage restraints have a significant impact on intra-euro area current account balances. Boing and Stadtmann(2016)use a similar research method when they empirically assess the impact of unit labor costs on current account balances in the Euro area. For this purpose,

(15)

15 they estimate a panel with annual observations from 2000 to 2013. Their findings suggest a

significant negative correlation between unit labor costs growth and the current account. However, they find that the effect of unit labor costs is mainly driven by productivity growth, and not so strong by wage growth. In addition, they included a dummy variable which takes a value of 1 from 2008 on. By doing this, they can research whether the effect of unit labor costs on the current account balance was different between the period before and the period after the financial crisis of 2008. Their results do not show a different negative effect of unit labor costs from 2008 on.

The effect that unit labor costs have on the current account balance can also be discussed via the real exchange rate channel. As discussed earlier in chapter 2.4, changes in unit labor costs have a strong correlation with changes in the overall price level (Kendrick, 1961; Ca'Zori and Schatz, 2007). This means than a decrease of unit labor costs will result in a real depreciation. The real deprecation will improve overall competitiveness and this will in turn result in an improvement of the current account balance.

Arghyrou and Chorteas (2008) find that after controlling for the role of income growth, a significant negative long-run relationship exists between the movements of the real exchange rate and the current account balance. They use the consumer price index for the euro zone countries from 1970 to 2005 and find that the two groups of countries with systematically

improving/deteriorating current account balances correspond to the two groups of countries that experience persistent real exchange rate depreciation/appreciation. Belke and Dreger (2011) draw the same conclusion: they find that the real exchange rate has a negative relationship with the

current account balance.

3.2 Arguments against a causal relationship from unit labor costs to the current account

balance.

Some new research on the rebalancing of the current account focuses on the role of domestic demand and imports as possible causes for divergence. Several authors advocate that the loss in competitiveness is a less important cause of current account deficits then domestic demand booms are (Diaz Sanchez and Varoudakis, 2013; Di Comite et al., 2013; Wyplosz, 2013; Communale and Hessel, 2014). In this literature, the increase in unit labor costs is seen as a side product of domestic demand growth, and not as the main cause of the current account deficit. The trends in domestic demand have recently gained more attention as possible drivers of the current account imbalances in the Euro area. Actually, a divergence in the strength of domestic demand growth occurred in the euro zone. This divergence took place between the peripheral member states and the economic stronger core countries, where growth was very strong in the former and much more subdued in the

(16)

16 latter. It is no coincidence that the divergence in the domestic demand growth follows the path of the divergence in current account balances. Indeed, the euro zone countries that saw a large increase in domestic demand also saw a substantially large deterioration of their current account balance.

One has to look at equation 3 to understand this explanation. Equation 3 states that economies which have more demand for investment or consumption than there are savings available in the home country have to finance the difference with capital from abroad. Economies which save little and invest much have thus low or even negative current account balances. When economies invest more then they save, we speak of a boom in domestic demand which

consequently increases the demand for labor. This increased demand for labor in turn causes wages to rise. This increase in wages together with low productivity growth causes unit labor costs to rise.

This suggests that the current account deficits could cause an increase in unit labor costs instead of the other way around. Wyplosz(2013) states that the current account deficits of euro zone countries prior to the European debt crisis were the consequence of excessive domestic demand in these countries. He states that bringing down demand will eliminate most of the factors associated with the crisis. This will namely improve the current account balance. Gaullier and Vicard (2012) agree and argue that current account imbalances are caused by divergences in domestic demand, instead of competitiveness problems.

The casual relationship between domestic demand and the current account balance has also been tested empirically. Diaz Sanchez and Varoudakis ( 2013) acknowledge that demand booms have been the key drivers of current account fluctuations. They estimate that changes in competitiveness have contributed less to current account imbalances then demand shocks. Furthermore, Communale and Hessel(2014) investigate the relative importance of price

competitiveness and domestic demand in explaining current account imbalances in a panel setup for Euro area countries. They conclude that competitiveness has a clear influence, but domestic demand booms have been more important than is often realized in the policy debate. They found that the influence of price competitiveness is most clear on export performance. On the other hand, domestic demand is by far the most important driver of imports. Their findings implicate that both improvements in competitiveness and decreases in demand can rebalance the current account balance, via an increase in exports and an decrease in imports respectively.

Although the literature is not unambiguous about the question in what way unit labor costs affect the current account balance, it is clear that they are related. As stated before, positive changes in competitiveness have positive effects on, among other things, export performance. Improving export performance will have a positive effect on the current account balance. The

(17)

17 conclusion is that the literature suggests that Greece can improve its current account by decreasing unit labor costs. This paper now continues to research empirically if the supposed relation between unit labor costs and the current account balance that has been suggested by the literature can be seen in Greece. Therefore, in the following chapter ,econometric analysis will be used in order to study this research question.

4. Empirical analysis

Zemanek et al. (2010) assess the importance of structural reforms and wage restraints for the current account balance in the Euro area. Boing and Stadtmann (2016) estimate a panel of Euro area countries from 2000 to 2013 and empirically assess the impact of unit labor costs on current account balances in the Euro area. They find that unit labor costs have a significant negative impact on the current account balance. In addition, they show that changes in unit labor costs are mainly driven by changes in productivity (see Eq. 2 for a definition of ULC). They make use of a panel of Euro area countries with annual observations. This paper, however, will make use of time series data and quarterly observations.

The original contribution of this paper consists of two aspects. First, this paper analyzes the effect that unit labor costs had on the current account balance. It does this for a specific country, Greece. Second, this paper makes a distinction between the period after the Greek structural reforms took place in 2010 and the period before the adjustment program was enacted and estimates the differences in the impact of unit labor costs between the two periods. Possibly, the European debt crisis has reinforced the need for ULC improvements and, hence, the effectiveness of these improvements on the current account balance is higher from 2010 on.

As was discussed in section 3, the literature is ambiguous about the question whether the causal relationship runs from unit labor costs to the current account balance or and vice versa. However, recent empirical research clearly shows that the relation runs from unit labor costs to the current account. Both Zemanek et al. (2010) and Boing and Stadtmann (2016) find this relation to be significant. This paper therefore assumes that the causal relation runs from unit labor costs to the current account balance and not vice versa. However, the hypothesis that the relation runs from the current account balance to unit labor costs will also be tested. This hypothesis will be tested by making use of a Granger causality test.

(18)

18

4.1 Empirical model

An OLS regression analysis, based on time series data, will be used to investigate the relationship between unit labor costs and the current account balance. The sample that is being used in this paper consists of 68 quarterly observations for Greece between the period 1999 and 2016. Quarterly data is being used because annual data does not provide sufficient observations. Large amounts of data make it easier to classify outliers and by definition, more observations result in a lower standard error2.

To estimate the impact of changes in unit labor costs on the current account balance an autoregressive distributed lag model will be used. This model is appropriate when using time series data (Stock and Watson, 2015). The autoregressive distributed lag model used in this paper will contain two lagged values of the percentage change in the current account balance and two lagged values of the percentage change in unit labor costs, which is common in this kind of literature (Zemanek et al., 2010). In addition, adding a second lag value substantially increases the 𝑟2 of the model, as can be seen in table 1. lets now turn the variables of the model. CA is the percentage change of the current account balance and CAlvl is the level of the current account balance. Moreover, ULC is percentage change in unit labor costs, GDP is the percentage change in gross domestic product and REER is the value of the real exchange rate. Furthermore, t is a time subscript. (Eq. 4) 𝐶𝐴𝑡 = 𝛽0+ 𝛽1𝑡𝑡+ 𝛽2𝐶𝐴𝑙𝑣𝑙𝑡−1+ 𝛽3𝐶𝐴𝑡−1+ 𝛽4𝐶𝐴𝑡−2+ 𝛽5𝑈𝐿𝐶𝑡+ 𝛽6𝑈𝐿𝐶𝑡−1

+ 𝛽7𝑈𝐿𝐶𝑡−2+ 𝛽8𝐺𝐷𝑃𝑡+ 𝛽9𝑅𝐸𝐸𝑅𝑡 + 𝛽10𝐴𝑆𝐸𝑡 + 𝑢𝑖

According to Stock and Watson (2015), many time series appear to be nonstationary. When time series are nonstationary the regression results can be biased or/and inefficient. To test whether the dependent variable CA has a stochastic trend and is thus nonstationary, an augmented Dickey-Fuller test is performed. This test is performed on a separate regression which can be found in table 3 in the appendix. The ADF t-statistic is the t-statistic testing the hypothesis that the coefficient on CAlvl_t−1 is zero (Stock and Watson 2015). The dependent variable CA is said to contain a stochastic trend when the null hypothesis is not rejected.

Similar to the dependent variable, the independent variables may also be of a nonstationary nature. However, using percentage changes for all the independent variables significantly lowers the chance of nonstationarity (Zemanek et al. , 2010). It is therefore standard in this type of literature to make use of percentage changes.

By including two lagged values of the main explanatory independent variable one can

2_𝑆𝐸

(19)

19 conduct a Granger causality test which tests if there is a significant relationship between the

dependent and the main independent variable (Granger, 1969). The null hypothesis states that the coefficient on the lagged values of ULC is equal to zero. When this hypothesis is rejected, it implies that ULC has predictive content for CA beyond that contained in the other regressors (Stock and Watson, 2015).

Hypothesis one, which states that changes in unit labor costs affected the current account balances in Greece, is corroborated if the coefficient β5 has a significant negative sign. What follows

is the second hypothesis, which states that Changes in unit labor costs affected the current account balance in Greece differently after the Greek adjustment program. The model has been extended to investigate this change:

(Eq. 5) 𝐶𝐴𝑡 = 𝛽0+ 𝛽1𝑡𝑡+ 𝛽2𝐶𝐴𝑙𝑣𝑙𝑡−1+ 𝛽3𝐶𝐴𝑡−1+ 𝛽4𝐶𝐴𝑡−2+ 𝛽5𝑈𝐿𝐶𝑡+

𝛽₆𝑈𝐿𝐶_𝑡−1+ 𝛽₇𝑈𝐿𝐶_𝑡−2+ 𝛽₈𝐺𝐷𝑃_𝑡+ 𝛽₉𝑅𝐸𝐸𝑅_𝑡 + 𝛽₁₀𝐴𝑆𝐸_𝑡+ 𝛽₁₁𝑅𝑚_𝑡 + 𝛽12𝑅𝑚𝑡∗ 𝑈𝐿𝐶𝑡 + 𝑢𝑖

A dummy variable has been created which equals zero in the period before the second quarter of 2010 and which equals 1 after the second quarter of 2010. It thus makes a distinction between the period before the structural reforms took place and the period after the structural reforms took place. This specification enables the testing of the question whether the relationship between CA and ULC is influenced by Rm, which marks the beginning of the structural reforms in Greece.

(20)

20

4.2 Data and variables

This research focuses on the current account balance of Greece and therefore the changes in the current account balance of Greece with the world in percent of GDP (𝐶𝐴) are being used as the dependent variable. This model specification is in line with the method used by Zemanek et al. (2010) and Boing and Stadtmann (2016).The growth rate of the nominal unit labor costs is the main explanatory variable that will be used. Nominal unit labor costs (𝑈𝐿𝐶) will be used instead of real unit labor costs because of two reasons. First, the European commission focused its policies at reducing the former, as opposite to the latter, to improve the current account in Greece (EC, 2012). It is of the opinion that the overall price level, which determines real unit labor costs, is much harder to influence then the costs of labor. The Greek government can namely directly influence total labor costs by various measures such as increasing overall competitiveness and decreasing the minimum wage. Although there is a strong correlation between increasing nominal unit labor costs and an increasing price level, the price level is much harder to influence (Kendrick, 1961). Since this paper aims to analyze the effects of the policies of the European commission, it will in line with the European commission use nominal unit labor costs rather than real labor costs as the main driver of the current account balance. Secondly, Zemanek et al. (2010) also makes use of nominal unit labor costs and since this paper has a similar research method compared to their research method, nominal unit labor costs are used in this paper also.

The real exchange rate (𝑅𝐸𝐸𝑅) is added because a strong domestic currency will influence the current account balance negatively. A strong Euro will namely make Greek products more expensive to non-Euro zone countries which inevitably lead to a lower demand for Greek products in those countries. This line of reasoning refers to the argument being made by Bennett and Zarnic (2008), who state that a strong Euro negatively affected Greek exports. A high real effective exchange rate will thus worsen the current account balance, because it makes Greek products relatively more expensive compared to non-Euro zone countries. But, since according to Kendrick (1961) and Ca’Zorzi and Schnatz (2007) the overall price level and unit labor costs are closely related, the real exchange rate and unit labor costs might be related as well. This might create a

multicolinearity problem. However, the correlation between the growth in unit labor costs and the level of the real exchange rate is 0.0496, which is far from significant.

The growth rate of GDP (𝐺𝐷𝑃) is added because it is expected that an economy which experiences a high GDP also tends to attract foreign investment (CPB, 2015). This will then worsen the current account balance, since the increase in investments is not matched with a similar increase in savings (see Eq. 3). In an empirical study, IMF (2013) find that the expected GDP growth has a significant and negative effect on the current account balance.

(21)

21 The Athens stock exchange (ASE) is added as a control variable. This variable is added to account for the possible effect that the financial crisis of 2008 might have had on the current account balance. Including this variable is common in this kind of literature (Frankel and Saravelos, 2010). However, it is not expected that ASE has a significant impact on CA.

Finally, 𝑅𝑚 will account for a possible structural break at the start of the Greek adjustment program. The dummy is coded as zero for the years between 1999 and the first quarter of 2010 and 1 for the period between the second quarter of 2010 and the second quarter of 2016. Furthermore, 𝑅𝑚 is multiplied by 𝑈𝐿𝐶, which results in 𝑅𝑚 ∗ 𝑈𝐿𝐶. This variable will measure the impact of 𝑈𝐿𝐶 on CA after the structural reforms took place. It can now be researched if changes in unit labor costs affect the current account balance in Greece differently after the Greek adjustment program. It is expected that 𝑅𝑚 ∗ 𝑈𝐿𝐶 will affect the 𝐶𝐴 negatively. The negative impact after the structural reforms might be larger than the negative impact that ULC has on CA for the whole period

examined. It can namely be the case that the structural reforms of 2010 made the channel, via which ULC is supposed to have an influence on CA, operate more efficiently. Indeed, Increasing

competitiveness and substantially deterioration labor laws were aimed at increasing the efficiency of the channel via which unit labor costs is supposed to affect the current account balance (EC, 2012). In addition, Bouthevillain and Dufrenot (2011) state that during times of crisis a fall in unit labor costs is accompanied by a decrease in labor demand. They estimated that the coefficient between unit labor costs and labor demand is 0.03. This would imply that a fall in unit labor costs will decrease wages, lowering unit labor costs even more. This would make Greek products more competitive, which in turn would lead to an improvement of the current account balance.

All the variables are thus presented as percentage changes. The exception is the

𝑅𝐸𝐸𝑅, which is reported as its absolute value and not as a percentage change. This method is in line with Boing and Stadtmann (2016). Because this paper looks at quarterly data, the change in the variables is from quarter to the previous quarter.

4.3 Estimation results

Test of hypothesis 1: Do changes in unit labor costs affect the current account balance in Greece? The estimation results related to the first hypothesis are reported in table 1. The regression containing one lagged variable for ULC and one lagged variable for CA has also been included in table 1. By doing this, it is possible to see the difference in estimation results between this regression and the regression being described in section 4.2.

(22)

22

The results of the first regression show that the relationship between the current account balance and unit labor costs is significantly negative. Also the second regression shows a significant negative relationship between the current account balance and unit labor costs. These results state that decreasing ULC will have a positive effect on CA. These estimations are in line with Zemanek et al. (2010) and Boing and Stadtmann (2016), who find similar results.

(23)

23 The lagged value(s) of ULC is/are tested to be different from zero in order to establish whether changes in ULC Granger-causes changes in the current account balance (see 4.2). The lags of ULC are only significantly different from zero in the second regression. Therefore, it is concluded that the changes in ULC Granger-cause changes in the current account balance. This means that the past values of ULC in the second regression appear to contain information that is useful for

forecasting changes in CA, beyond that contained in past values of changes in CA.

The estimated coefficients of GDP and REER are in accordance with theory and, thus, corroborate the robustness of the estimation results. The relationship between the current account balance and the REER is negative. This conforms the statement that a strong Euro negatively effects the current account balance. However, the negative impact becomes less significant in the second regression. The coefficient of GDP growth is negative and affects the current account balance thus negatively. This confirms that higher economic growth worsens the current account balance.

To see whether CA has a stochastic trend, the one lagged variable CAlvlt−1 is tested to be

zero (see section 4.2). The result can be seen in table 3. The hypothesis that CA contains a stochastic trend cannot be rejected at the 10% significance level.

The literature is ambiguous about the question whether the causal relationship runs from unit labor costs to the current account balance or and vice versa. Therefore, a Granger causality test has been included which tests the hypothesis that the causal relationship runs from unit labor costs to the current account balance. The outcome of this test is presented in table 5. It is concluded that the past values of CA indeed contain information that is useful for forecasting changes in ULC.

Test of hypothesis 2: Changes in unit labor costs affected the current account balance in Greece differently after the Greek adjustment program.

Table 2 shows whether the effect of ULC on CA changes after the structural reforms in Greece took place. For this purpose, a dummy was defined which takes a value of 1 from the third quarter of 2010 on. Again, table 2 contains two regressions which differ from each other in the number of lags they contain. The first regression contains 1 lag of both ULC and CA, the second regression contains 2 lags of both the former and the latter.

(24)

24

(25)

25 Both the first and the second regression show that Reform * ULC have a significant impact on the current account balance. It can thus be stated that a decrease in ULC from the second quarter of 2010 on resulted in a significant increase of the current account balance. One interesting result is that the negative effect of ULC on the whole period examined is not significant in both regressions. This means that ULC did not have a significant impact in the years between the first quarter of 1995 and the second quarter of 2010, whereas it had a significant impact on CA from the second quarter of 2010 on. This result is different from the result of Boing and Stadtmann (2016), who did not find any significant difference in impact of unit labor costs on the current account balance between the periods they examined. However, it is worth noting that they used the financial crisis as a dummy, whereas this paper assumes the Greek adjustment program to be the turning point. This paper also accounts for the effect of the economic crisis on the current account balance. But, it does this by adding the variable ASE to the regression, and not be creating a time dummy.

The change in effect of unit labor costs after the structural reforms in Greece might be due to the fact that the European debt crisis has reinforced the need for ULC improvements in Greece and, hence, their effectiveness is higher from the second quarter of 2010 on. An explanation for this different impact can be found in section 2.3, where the structural reforms that were included in the adjustment program are discussed. As was discussed earlier, ULC have an impact on the CA via the price channel, where a decrease in ULC is supposed to bring about a decrease in prices. This will eventually make Greek products more competitive compared to economies which did not saw their ULC decline (Alexiou, 2013). This channel, which runs from ULC to the overall price level, must thus have worked more efficiently after the structural reforms took place. Indeed, the adjustment program contained measures to promote the efficiency of this channel. The Greek government namely took measures to promote labor market flexibility and also lowered the minimum wage (EC, 2012). These measures promoted downward pressure on wages, and thus on unit labor costs. As was discussed in section 2.3, these measures caused ULC to decline relatively sharp, which in turn caused inflation to decline. Myant et al. (2016) indeed state that nominal wage cost decreases in Greece from 2011 onwards were passed on to exported goods, and thus decreased the prices of these goods. They also argue that this effect was larger in the period after 2011 then in the period before 2011, probably due to stronger competition in international markets (Myant et al., 2016). This namely may have caused prices to adjust to changes in wages swifter, because not doing so may have resulted in pricing themselves out of the market.

Nevertheless, a detailed discussing about the differences in economic structure between two periods lies beyond the scope of this paper. However, the author of this paper recognizes the contribution that such a discussion can have and therefore advises that the topic be researched

(26)

26 further.

The effect of GDP and REER remained unchanged compared to table 1. The effect of GDP remains negative, but is not significant anymore. The relationship between the current account balance and the REER also remains negative, corroborating the robustness of the estimation results.

5. Conclusion

A sharp increase in unit labor costs in Greece resulted in a deterioration of the competitive position of their economy. This led to an unsustainable current account deficit. The European countries and the IMF therefore designed an adjustment program which was aimed at restoring competitiveness. Structural reforms that promoted downward pressure on wages and which thus decreased unit labor costs were taken. These measures were supposed to improve the current account balance. Firstly, the goal of this paper therefore was to investigate whether or not changes in unit labor costs is one of the drivers of the account balance in Greece. In addition, this paper also looked if the impact that unit labor costs had on the current account balance was different in the period after the structural reforms compared to the period before these reforms took place.

The main argument supporting the causal relationship between unit labor costs and the current account balance states that decreasing unit labor costs reduces the prices of its products. This would make Greek products more competitive on the international market. This in turn would lead to an increase in exports and thus an improvement of the current account balance.

The main argument against the above relationship is the claim that the causal link goes from changes in the current account balance to unit labor costs, and not vice versa. Here, the increase in unit labor costs is seen as a side product of domestic demand growth, and not as the main cause of the current account deficit. The increase in domestic demand namely increased the demand for labor, putting upward pressure on wages.

An OLS regression was used to investigate the relationship between unit labor costs and the current account balance. The outcome of this research supports the hypothesis that unit labor costs have affected the current account balance in Greece negatively. A decrease in unit labor costs will thus lead to an improvement of the current account balance. This finding has strong policy implications, since it confirms that the actions of the combined European commission and IMF regarding Greece proved to be successful. However, since there could be reversed causality, the findings of this paper do not prove that the opposing hypothesis is not true. In fact, the Granger causality test indeed shows that the past values of CA contain information that is useful for forecasting changes in ULC. It is therefore recommended that it is researched further which causal

(27)

27 direction is more relevant, since both are true.

This research can be further improved by including the change in unit labor costs of the different sectors of the economy in the regression. By doing this, the effect that ULC has on CA can be separated between these sectors. It can then be seen in which sector a change in unit labor costs has the biggest effect on the current account balance. Policymaker can then direct their policy to the sector in which it is most effective. However, the available data did not allow for such a distinction to be made.

Another outcome of this research is that changes in unit labor costs affected the current account balance in Greece differently after the structural reforms in 2010 took place, compared to the period before 2010. This might be due to the fact that the European debt crisis has reinforced the need for ULC improvements in Greece and, hence, their effectiveness is higher from the second quarter of 2010 on. There are two explanations for the increased effectiveness of ULC on CA in times of economic crisis. The first explanation states that stronger competition in international markets were a cause. This namely may have caused prices to adjust to changes in wages swifter, because not doing so may have resulted in pricing themselves out of the market. The second explanation states that during times of crisis a fall in unit labor costs is accompanied by a decrease in labor demand. This would result in an even larger drop in unit labor costs, reinforcing the effect of the original decrease in unit labor costs on the current account balance. A more detailed research on the different effect of ULC on CA during a economic crisis is desirable, important indeed. It might induce policy makers to alter their approach regarding unit labor costs, when the economy in question finds itself stuck in a crisis.

(28)

28

References

Alexiou, C., Nellis, J.C. (2013) Challenging the raison d’etre of internal devaluation in the context of the Greek economy, Panoeconomicus, 60 (6), 813–836.

Arghyrou, M., Chortareas, G. (2008), Current Account Imbalances and Real Exchange Rates in the Euro Area, Review of International Economics, vol. 9, issue 5, pp. 747-764

Belke, A., Dreger, C. (2011), Current account imbalances in the euro area: Catching up or competitiveness?, DIW Berlin, Discussion Papers, no. 1106.

Bennett, H.Z., Zarnic, Z. (2008),International Competitiveness of the Mediterranean Quartet: A Heterogeneous-Product Approach. IMF Working Paper 204. International Monetary Fund.

Bitzis, J., Paleologos, M., Papazoglou C. (2008), The Determinants of the Greek Current Account Deficit: The EMU Experience, Journal of International and Global Economic Studies, 1 (1), June 2008, 105-122.

Blanchard, O.J., Milesi-Ferretti. (2011), (Why) Should Current Account Imbalances Be Reduced? IMF Staff Discussion Note SDN/11/03.

Bouthevillian, C., Dufrenot, C. (2011), Are the Effects of Fiscal Changes Different in Times of Crisis and Non Crisis? The French Case, Revue D’Economie Politique 121, 271-407.

Ca’Zorzi, M., Schnatz, B. (2007), Explaining and Forecasting Euro Area Exports: Which Competitiveness indicator performs best? ECB, Working Paper Series, no. 833.

Calvo, G.A., Reinhart, C. (1999), When Capital Inflows Come to a Sudden Stop: Consequences and Policy Options, unpublished manuscript, University of Maryland at College Park.

Catão, L.,Milesi-Ferretti. (2011), External Liabilities and Crises. Manuscript, International Monetary Fund.

Chinn, M.D., Prasad, E.S. (2000). Medium-Term Determinants of CurrentAccounts in Industrial and Developing Countries: An Empirical Exploration, IMFWorking Paper 46, International Monetary Fund.

Comunale, M., Hessel, J. (2014),Current Account Imbalances in the Euro Area: Competitiveness or Financial Cycle? DNB Working paper No. 443, pp. 1–46.

(29)

29 Darvas, Z. (2012) Internal adjustment of the real exchange rate: Does it work?

http://www.voxeu.org/article/internal-adjustment-real-exchange-rate-does- it-work

De Grauwe, P. (2012) In search of symmetry in the Eurozone. Centre for European Policy Studies. CEPS Policy Brief No. 268

European Commission (2010), The Economic Adjustment Programme for Greece. European Economy Occasional Paper 61, Brussels.

European Commission (2012), The Second Economic Adjustment Programme for Greece. European Economy Occasional Paper 94, Brussels.

Frankel, J.A., Saravelos, G. (2010), Are Leading Indicators of Financial Crises

Useful for Assessing Country Vulnerability? Evidence from the 2008-09 Global Crisis. Working Paper 16047, National Bureau of Economic Research.

Gaulier, G., Vicard, V. (2012), Current account imbalances in the euro area: competitiveness or demand shock?Banque de France, Quarterly Selection of Articles 27, pp. 5-26.

Golub, S., Hseih, C. (2000), Classical Ricardian Theory of Comparative Advantage Revisited, Review of International Economics, Vol. 8, Issue 2, pp. 221-234

Gourinchas, P.O., Obstfeld, M.(2012), Stories of the Twentieth Century for the Twenty-First.American Economic Journal: Macroeconomics4 (January 2012): 226-265.

Granger, C. W. J. (1969), Investigating Causal Relations by Econometric Models Cross-Spectral Methods. Econometrica 37.

Gruber, J.W., Kamin, S. B. (2005), Explaining the Global Pattern of Current Account Imbalances, International Finance Discussion Papers 2005-846, The Federal Reserve Board.

Huh., Guk, C., Trehan, B.(1992), Modelling the Time Series Behavior of the Aggregate Wage Rate,Working Paper 92–04. San Francisco: Federal Reserve Bank of San Francisco.

Hume, D. (1752),Political Discourses. Edinburgh: Printed by R. Fleming, for A. Kincaid and A. Donaldson, 1752.

(30)

30 Kendrick, J.W. (1961),Relative Changes in Productivity, Prices, and Resource. In J. W. Kendrick,

Productivity Trends in the United States (pp. 189 - 228). Princeton: Princeton University Press.

Kennedy, M., Sløk, T. (2005), Structural Policy Reforms and External Imbalances, OECD Economics Department Working Papers, No. 415, OECD Publishing, Paris.

Krugman, P., Obstfeld, M. (2015), International Economics: Theory & Policy, Tenth Edition, Global Edition, Pearson International, Essex, England

Lipschitz, L., McDonald, D. (1992) Real Exchange Rates and Competitiveness, Austrian Econ Pap 19: 37-69.

Mendoza, Enrique, G., Terrones, M. E. (2008), An Anatomy of Credit Booms: Evidence from Macro Aggregates and Micro Data. Working Paper 14049.

Monokroussos, P., Thomakos, D. (2012), A technical study on the determinants of Greece’s current account position, Eurobank EFG, Economy & Markets, Vol. VII, No. 2, april.

Myant, M., Theodoropoulou, S., Piasna, A. (2016), Unemployment, internal devaluation and labor market deregulation in Europe. European Trade Union Institute (ETUI), Brussels.

Obstfeld, M.,Rogoff, K. (2010), Global Imbalances and the Financial Crisis: Products of Common Causes. San Francisco: Federal Reserve Bank of San Francisco.

Obstfeld, M. (2012b), Does the Current Account Still Matter? American Economic Review, 102(3), 1-23.

Olthetene, E.A . (2003), Greece in the European Union: Policy Lessons from Two Decades of Membership.The Quarterly Review of Economics and Finance, Vol. 43, Issue 5, 2003, pp. 774-806.

Papademos, (2007) Inflation and competitiveness divergences in the Euro area countries. [speech] Retrieved from https://www.ecb.europa.eu/press/key/date/2007/html/sp070907_2.en.html.

Philip, R.L., Maria, G., Milesi-Ferretti. (2011),External Adjustment and the Global Crisis. IMF Working Paper WP/11/197.

Pisani-Ferry, J. (2012a), The Euro Crisis and the New Impossible Trinity. Bruegel Policy Contributions, 1.

(31)

31 Price, W.H. (1905), “The Origin of the Phrase ‘Balance of Trade.” Quarterly Journal of Economics, Vol 20, p. 157-67.

Reinhart, C.M., Reinhart, V.R. (2009),Capital Flow Bonanzas: An Encompassing View of the Past and Present. In Jeffrey A. Frankel and Christopher Pissarides, eds., International Seminar on

Macroeconomics 2008. Chicago: University of Chicago Press.

Ricardo, D. (1817), On the Principles of Political Economy and Taxation, Chapter IV, John Murray, First edition, retrieved from Google Books

Rojas-Romagosa, O.C. (2015),Literature survey on the theoretical explanations and empirical determinants of current account balances. Den Haag: CPB.

Schularick, M., Taylor, A.M. (2012) Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870-2008. American Economic Review 102.

Shambaugh, J.A.(2012), The Euro’s Three Crisis, Brookings Papers on Economic Activity. P. 157-211.

Stiglitz, J.E. (2016). The Euro: how a common currency threatens the future of Europe. New York: W. W. Norton & Company.

Stock, J., Watson, M. (2012), Introduction to Econometrics, Updated Third Edition, Pearson Educated Limited, 2015.

The European Central Bank, Statistics glossary. (2017). Unit labour costs. Retrieved from http://www.ecb.europa.eu/home/glossary/html/act2u.en.html#116

Van Ark, T. (2002), Capital Formation and Productivity Growth in South Korea and Taiwan: Beating Diminishing Returns through Realising the Catch-Up Potential. University of Groningen.

Wyplosz, C. (2013), Eurozone Crisis: It’s About Demand, not Competitiveness, The Graduate Institute, Geneva, and CEPR.

Zemanek, H., Belke, A., & Schnabl, G. (2010, Vol.7). Current account balances and structural adjustment in the euro area. International Economics and Economic Policy , p. 83-127.

(32)

32

Appendix 1

Table 4: Large-sample critical values of the augmented Dickey-Fuller statistic

Deterministic regressors 10% 5% 1%

Intercept only

-2.57 -2.86 -3.43 Intercept and time trend

-3.12 -3.41 -3.96

Source: Stock and Watson, 2015

(33)