Discussion Paper
No. 2016-30 | June 29, 2016
Oscar Bajo-Rubio, Burcu Berke and Vicente Esteve
The Effects of Competitiveness on Trade Balance: The Case of Southern Europe
(Published in Recent Developments in Applied Economics)

Abstract

According to the conventional wisdom, “peripheral” Southern European countries members of the euro area (Greece, Italy, Portugal, and Spain) suffer from a problem of competitiveness. Since devaluation is not possible because they are part of the euro, the adjustment should come through decreasing wages and prices in these countries, which, by improving the trade balance, should lead to recover the previous levels of employment and growth. In this paper, the authors estimate trade balance equations for the Southern European countries, both for total trade and for the trade performed with the European Union; and taking three alternative measures of the real exchange rate, i.e., based on consumption price indices, export prices and unit labour costs, respectively. Their main conclusion is that demand seems to be more relevant than relative prices when explaining the evolution of the trade balance.

JEL Classification:

F31, F41, F45

Links

Cite As

[Please cite the corresponding journal article] Oscar Bajo-Rubio, Burcu Berke, and Vicente Esteve (2016). The Effects of Competitiveness on Trade Balance: The Case of Southern Europe. Economics Discussion Papers, No 2016-30, Kiel Institute for the World Economy. http://www.economics-ejournal.org/economics/discussionpapers/2016-30


Comments and Questions



Ileana Tache, Transilvania University of Brasov, Romania - Invited Reader Comments
July 11, 2016 - 08:36
This is a paper with a consistent look, given both by the theoretical background and the empirical part – a rigorous estimation of equations for trade balance. The paper challenges the orthodox theory of international trade and sheds light on the determinants of trade flows: the classical solutions of lowering relative prices (when devaluation is not possible – the case of the analyzed countries) in order to restore export competitiveness are not valid anymore. Arriving at the conclusion that external demand is much more important than the level of internal unit labor costs means that, in the context of the globalization phenomenon, the external factors become extremely significant. What does it mean, eventually, the term “domestic” in a more and more globalized world? Moreover, in a very complex and highly diversified world economy, with a sleepy growing sophisticated demand, it seems that the price level is no more the main factor affecting competitiveness. The empirical results of this paper bring solid arguments for this obvious tendency. The main result of the paper – that the trade balance is not clearly related to changes in the real effective exchange rate – was obtained by studying the case of Southern Europe (Greece, Italy, Portugal and Spain). It would be interesting to check the validity of the respective conclusion for other EU countries.

Oscar Bajo-Rubio - answer to comments
September 19, 2016 - 03:28
Many thanks for your very generous comments. We detail here the changes introduced in the new version. First, we analyse the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced are:• We justify the choice of the sample period.• Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity. Finally, we provide an extended discussion of the results:• We mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country.• We stress the role of some other factors affecting competitiveness beyond prices, following Myro (2015).• We provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990).

Anonymous - Invited Reader Comment
July 26, 2016 - 08:22
see attached file

Oscar Bajo-Rubio - answer to comments
September 19, 2016 - 03:36
Many thanks for your very interesting comments, which have made us to think additionally on our paper. Regarding your first point, we have preferred to give a qualitative, rather than quantitative, flavour to our results, given the limitations of any econometric method and the precaution with which any quantitative estimate should be taken. On the other hand, introducing in the equation a proxy for product differentiation is no doubt an interesting avenue of research, but beyond the scope of this paper: we are interested in estimating a basic demand equation for the trade balance, i.e., in terms of (foreign and domestic) output and the real exchange rate. We have preferred instead to extend the paper in the revised version by analysing the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced in the revised version are:• We justify the choice of the sample period.• Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity: this is your point 5. Finally, we provide an extended discussion of the results:• We mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country: this is your point 3.• We stress the role of some other factors affecting competitiveness beyond prices, following Myro (2015).• We provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990).

Anonymous - Referee Report 1
August 01, 2016 - 08:42
see attached file

Oscar Bajo-Rubio - answer to referee 1
September 19, 2016 - 03:34
Many thanks for your very interesting comments. Regarding your first point, we have some results for trade with the Eurozone, but we have preferred not include them to not making the paper unnecessarily long. The same applies to your last point. The reason for the different sample periods for total trade and trade with the EU is, of course, data availability; we mention this explicitly in p. 7 of the revised version (your point 2). The main change in the revised version relates to your point 3. Specifically, we analyse the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced are as follows. Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity. Also, we provide an extended discussion of the results:• We mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country.• We stress the role of some other factors affecting competitiveness beyond prices, following Myro (2015).• We provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990).

Rafael Myro - Referee Report 2
August 01, 2016 - 08:43
"The effects of competitiveness on trade balance: The case of Southern Europe"Review by Rafael MyroThis paper contents a brilliant contribution to the debate on competitiveness. Through the formulation of a dynamic model of trade balance, the authors show that the external equilibrium is more guided by quantities than prices, according with the result from other authors well quoted in the text. So, an external equilibrium is a well measure of an excess of demand than a measure of lacking competitiveness, giving support to the discussion about this matter made by various authors and by Myro (2015) in a reference to Spain. Beside in some of the estimates, the ULC are not significant at all in the long run, in accordance with some other estimates for the countries contemplated.The paper is clear, with the hypothesis well settled, the theoretical model well defined and founded and the econometric work very robust. Its conclusion are fundamental for better capture the competitiveness matters and the true dynamic of external equilibrium.So I have no suggestion to do and from my personal view, the paper can be published.Myro, R. (2015): España en la economía global. Claves del éxito de las exportaciones españolas. RBA, Barcelona.

Oscar Bajo-Rubio - answer to referee 2
September 19, 2016 - 03:33
Many thanks for your very generous comments. We detail here the changes introduced in the new version. First, we analyse the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced are:• We justify the choice of the sample period.• Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity. Finally, we provide an extended discussion of the results:• We mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country.• We stress the role of some other factors affecting competitiveness beyond prices, following Myro (2015).• We provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990).

Antonio Gómez Gómez-Plana - On Bajo-Rubio, Berke and Esteve
August 02, 2016 - 09:38
This paper studies the relationship between trade balance and a set of variables (i.e., income and real exchange rate under several definitions). The literature has been inconclusive in some aspects and the paper provides new insights. Overall, the paper presents a relevant question, with an adequate methodological approach and a relevant selection of the case. The conclusions confirm also the difficulties to reach a general pattern across countries. One of the main points is the role of the definition of real exchange rate. This is a common problem in empirical models of international trade, so the paper reinforces the requirement of testing this definition. Also the length of the time series seems a crucial point, as authors explain in the paper. The paper deserves its publication.

Oscar Bajo-Rubio - answer to comments
September 19, 2016 - 03:29
Many thanks for your very generous comments. We detail here the changes introduced in the new version. First, we analyse the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced are:• We justify the choice of the sample period.• Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity. Finally, we provide an extended discussion of the results:• We mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country.• We stress the role of some other factors affecting competitiveness beyond prices, following Myro (2015).• We provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990).

Marek Kosny - Invited Reader Comment
August 08, 2016 - 08:12
The article concerns very interesting and up to date issue of expected impact of real effective exchange rates (and related external competitiveness of economy) on trade balance. In this sense it can be treated as a voice in a discussion on the legitimacy of the suggested by IMF, World Bank and EU ways to achieve stability of the economies of the southern European countries.The results are well embedded in the theory and very interesting. Though, in my opinion it would be worthwhile to distinguish different groups of goods (related to sectors of the economy) in the analysis. The authors suggests (in Conclusions) that some big companies "enjoy productivity levels above the average of the economy". It would be very interesting to identify characteristic of these companies – are they oriented to export, are they dominated by foreign capital and so on.The analysis can be also extended by introducing an additional variable, reflecting impact of organized forms of export promotion, which vary widely between countries.

Oscar Bajo-Rubio - answer to comments
September 19, 2016 - 03:30
Many thanks for your very generous comments. We detail here the changes introduced in the new version. First, we analyse the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced are:• We justify the choice of the sample period.• Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity. Finally, we provide an extended discussion of the results:• We mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country.• We stress the role of some other factors affecting competitiveness beyond prices, following Myro (2015).• We provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990).

Jorge Uxó - Invited Reader Comment
August 09, 2016 - 09:54
see attached file

Oscar Bajo-Rubio - answer to comments
September 19, 2016 - 03:32
Many thanks for your very generous and insightful comments. We detail here the changes introduced in the new version. Regarding your specific comments, in the revised version we try to stress more strongly our main results in the conclusions; in particular, we provide a connection of our results to the concept of a wage-led regime due to Bhaduri and Marglin (1990). Also, we mention the four “critical channels” of Ottaviano et al. (2007) that can allow to improve the competitiveness of a country, and stress the role of some other factors affecting competitiveness beyond prices following Myro (2015). On the other hand, we analyse the possible disconnection between exchange rates and trade due to the expansion of global value chains, which could weaken the relationship between exchange rates and trade. The reason is that now an exchange rate depreciation would improve the competitiveness of just a fraction of the value of final exports. In particular, we first analyse whether the estimated equation is unstable by means of the tests proposed by Kejriwal and Perron (2008, 2010). Next, we test for the presence of cointegration with structural changes following Arai and Kurozumi (2007). In addition, our econometric equation is re-estimated by subperiods in order to see if (i) the size and significance of the estimated coefficients have changed, and (ii) there is still cointegration. Other changes we have introduced are:• We justify the choice of the sample period.• Together with the Phillips-Perron unit root tests, we present now the KPSS tests on the null hypothesis of stationarity.