Recent French Export Performance: Is There a Competitiveness Problem?
Author: Alain N. Kabundi1 and Mr. Francisco d Nadal De Simone
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Contributor Notes

Author’s E-Mail Address: akabundi@uj.ac.za ; francisco.nadaldesimone@bcl.lu
Publication Date:
01 Jan 2009
eISBN:
9781451871494
Language:
English
Keywords:
WP; demand shock; export; SITC; France
Download PDF (1.8 MB)

Recently, the export performance of France relative to its own past and relative to a major trading partner, Germany, deteriorated. That deterioration seems related to the geographical destination and product composition of trend exports. Faced with an increase in unit labor costs or in its terms of trade, France adjusts relatively less via price and wage changes, and more via employment changes. Given that SMIC convergence resulted in a significant increase in unit labor costs, foreign sector difficulties might be structural. Trade flows relevance and euro area policy constraints highlight the importance of structural reforms that increase markets flexibility.

Abstract

Recently, the export performance of France relative to its own past and relative to a major trading partner, Germany, deteriorated. That deterioration seems related to the geographical destination and product composition of trend exports. Faced with an increase in unit labor costs or in its terms of trade, France adjusts relatively less via price and wage changes, and more via employment changes. Given that SMIC convergence resulted in a significant increase in unit labor costs, foreign sector difficulties might be structural. Trade flows relevance and euro area policy constraints highlight the importance of structural reforms that increase markets flexibility.

I. Introduction

Evidence shows that a significant part of French activity fluctuations has a foreign source (e.g., Kose, Otrok, and Whiteman, 2003; and Kabundi and Nadal De Simone, 2007). In particular, since 2000, French foreign sector performance has experienced a substantial deterioration vis-à-vis its own past and relative to Germany. Some observers have suggested that the country has not benefited fully from the opportunities offered by the rapid economic performance of emerging Asian economies and the eastward expansion of the European Union (EU). Therefore, the question arises as to whether France is suffering from a competitiveness problem. This question has had, so far, an elusive answer. Traditional variables that explain international trade, such as the exchange rate, relative unit labor costs, and demand pressure seem insufficient to illuminate the recent decline in France’s export performance. Residuals from econometrically-estimated equations indicate a substantial drag on exports since 2001, not attributable to the standard global demand and price/cost factors.3

In addition, equilibrium exchange rate analysis indicates that France’s real effective exchange rate is largely in line with fundamentals. National account data show, however, that changes in export margins have cushioned the effects of the euro fluctuations. Cost competitiveness of French producers worsened in 2005 and early 2006, though it remains in line with its long-term average. Despite that producers lowered export prices in euros to maintain price competitiveness, the external position deteriorated during the period.

Hence, the relative underperformance of exports in past years may point to structural factors that leave French firms behind the global expansion. A more flexible economy should be able to reorient the destination of its exports and product mix toward fast-growing economies and sectors. Indeed, a sectoral study of total factor productivity (TFP) growth in manufacturing found that, while France does not lag significantly behind the United States in terms of level, TFP growth is hampered by the high ratio of minimum to median wages.4 Staff analysis also suggests that as France has become more sensitive to the global economy over time, it has tended to adjust more through changes in employment and productivity than through wage flexibility, strengthening the case for more structural reforms.5

This paper performs a descriptive analysis of French export data by destination and by SITC product classification distinguishing between the cyclical and the trend components of the series. Next, it analyzes the behavior of prices and quantities following a domestic and a foreign shock to the French economy; the paper contrasts and compares the reaction of French and German variables to shocks to unit labor costs and to terms of trade.

Globalization has greatly influenced economies over the past three decades. Countries’ boundaries have dropped through intensive trade of goods and services, and financial integration. Economies have benefited from trade and foreign direct investment (FDI). Conversely, globalization can make countries more vulnerable to external shocks. Crises can be severe and contagion can spread rapidly to other parts of the globe, as recently exemplified by the subprime crisis that started in the United States.

There is a consensus in the literature that globalization has largely positive effects. Globalization fosters comovement of macroeconomic variables across countries through trade integration and financial market integration (IMF, 2001; and Imbs, 2004). An increase in exports in one country boosts economic activity of the recipient country (Canova and Dellas, 1993; and IMF, 2001). Such spillover effects lead to high correlation of business cycles.

The integration of financial markets has also contributed to the synchronization of business cycles through the opening of countries’ capital accounts. Financial prices have become more synchronized through arbitrage (e.g., the global slowdown of 2000–01 was caused by the crash of the technology stock market in the United States). Financial comovements tend to be substantially larger than comovements in the real economy, and financial comovement has increased for financial markets over time (Brooks, Forbes, and Mody, 2003).

On the empirical front, most findings show increasing synchronization of economic variables across countries. Kose, Otrok, and Whiteman (2005), using the Bayesian dynamic factor model, extract common components in output, consumption, and investment and find that the degree of synchronization of business cycles of major macroeconomic aggregates across G-7 countries has increased over time. On the nature of shocks that drive the comovement, they find that oil-price shocks are behind synchronization of cycles during the “common shocks” period. Nadal-De Simone (2002), using a concordance index proposed by Harding and Pagan (2002) and the dynamic factor model of Stock and Watson (1991), finds evidence of a global component as well as a regional component that explains the comovement between European economies themselves and with the United States. In the same vein, Monfort and others (2004); Kose, Otrok, and Whiteman (2003); Malek Mansour (2003); Yang (2003); Lumsdaine and Prasad (2003); Bordo and Helbling (2004); and Canova, Ciccarelli, and Ortega (2007) support the view that fluctuations of most macroeconomic variables across developed countries are mainly driven by a global factor.

In contrast, Kose and Yi (2006); Kose, Prasad, and Terrones (2003); Stock and Watson (2003); and Heitz, Hild, and Monfort (2004) find that G-7 business cycles have become less synchronized. A possible reason is that trade flows could lead to increased specialization resulting in changes in the nature of business cycles. Trade ties are closely related to a rise in inter-industry specialization across nations, and then industry-specific shocks are the main driving forces of business cycles. Synchronization may be thus reduced. Similarly, international financial linkages could also stimulate production through the reallocation of capital in a manner consistent with countries’ comparative advantage (Imbs, 2004), which in turn reduces business cycle synchronization.

Other studies have emphasized the sources of shocks, their spillovers, and channels of their transmission. Recent examples include the study of the monetary transmission mechanism in the euro area using structural VAR analysis by Peersman (2005); Canova, Ciccarelli, and Ortega (2007); and Ciccarelli and Rebucci (2006). Similarly, Canova and Ciccarelli (2006), using a VAR with time-varying parameters, find a positive and significant effect of U.S. GDP growth shock on France and Italy, but a negligible effect on German GDP growth. Canova (2005) uses a structural VAR approach and finds that U.S. monetary shocks have a strong influence, while real supply and demand shocks have a minor effect. Given the limitations of the VAR methodology—the most conspicuous being that it cannot accommodate a large panel of series without the risk of running short of degrees of freedom—Stock and Watson (1998 and 2002) use the approximate structural dynamic factor model on a large panel of developed countries’ variables. Like Kabundi and Nadal De Simone (2007) and Eickmeier (2007), find a positive and significant effect of U.S. demand shocks on French and German output, while EU supply shocks tend to have important effects on French and German output.

The high degree of integration, and with it the exposure of countries to shocks, stresses the importance of good and factor markets flexibility. Economies’ flexibility to absorb domestic-and foreign origin shocks takes paramount importance, even more so when countries’ policy menu is restricted in some sense such as by participation in a currency area. Not surprisingly, competitiveness issues have been taken to the front line of the economic and political debate.

This study contains several findings. (1) Divergences in recent trade performance between France and Germany are not related to the cyclical part of trade but to its trend. (2) For most categories of products, France’s export cyclical component is less volatile than Germany’s. (3) In the 2000s, France’s trend export growth rate while higher than in the 1990s, was less than 60 percent Germany’s. (4) Both France and Germany faced a negative common factor in the 2000s, most likely due to the euro appreciation. (5) However, idiosyncratic factors were negative on average for France and positive for Germany. (6) The French economy seems less flexible to adjust to a negative shock to unit labor costs in manufacturing or to its terms of trade: the adjustment tends to be done relatively more via quantities than via prices suggesting the need to make labor and product markets as flexible as possible.

Section II discusses the data and elaborates on the methodology to deal with non-stationarity in the data. Section III describes the cyclical and trend components of French exports by destination and by product. Section IV analyzes the response of the French economy to a shock to unit labor costs in manufacturing and to the terms of trade. Section V discusses the policy implications of the study.

II. Data and Non-Stationarity

A. Data

This study uses two large data panels. The first one comprises 396 quarterly macroeconomic series and 106 Direction of Trade (DOT) series of trade by country (for a total of series N = 502). DOT series include imports and exports to the euro area, the EU, Ascension countries, Canada, the United States, the United Kingdom, Japan, China, Asia, Latin America, and the rest of the world. The second data panel contains 396 quarterly macroeconomic series and 110 series of trade by SITC Revision 3 category of products (for a total of series N = 506). The sample period is 1981:Q1–2006:Q4, or 104 observations for the two data panels (i.e., T = 104). The countries included in the sample are France, Germany, Japan, the Netherlands, the United Kingdom, and the United States. In addition to national variables, a set of global variables is included, containing such items as crude oil prices, a commodity industrial inputs price index, world demand, and world reserves. The variables cover the real sector of the economy including consumption, investment, international trade in goods and services, portfolio flows and FDI flows, prices, financial variables, and confidence indicators. All variables have had their seasonal component removed. The complete list of variables used in this study is in Annex I.

B. Dealing with Non-Stationarity

For estimation purposes, series have to be covariance stationary. Instead of applying unit root tests to determine the degree of integration of the series and then difference or detrend them depending on whether they are I(1) or I(0) with a deterministic trend, respectively, the Corbae-Ouliaris Ideal Band-Pass Filter was used. See Appendix I for a technical description of the filter. The reason for this approach is twofold. First, as is well known, currently available unit root tests have low power and often the decision on the degree of integration of the series has to be based on subjective judgment. Second, it is also known that first differencing removes a significant part of the variance of economic time series. Third, the ideal band-pass filter of Corbae and Ouliaris is consistent, is not subject to end-point problems and has no finite sampling error. As an illustration of these points, note the large share of variance that first differencing of French real GDP produces at the business cycle frequency band (between 6 and 32 quarters, according to the NBER definition of business cycles) (text figure).

UF1

France: Spectra of Real GDP Filtered

(Y axis: spectrum; X axis: periodicities in quarters)

Citation: IMF Working Papers 2009, 002; 10.5089/9781451871494.001.A001

III. Descriptive Part: Facts Without the Noise

Several interesting features of recent French export performance are clear from the data once the noise of short-term fluctuations is removed. First, the cyclical components of exports by country and products of France and Germany follow the same pattern, mimicking quite closely their business cycles (Figure 1). Export cycles of both countries portray a picture of negative growth in the early 1980s and 1990s, and at the end of the 1990s. The U.S. driven early-1980s recession, the European 1993 recession and the end of the stock market “bubble” at the end of the 1990s are clearly correlated to exports behavior (IMF, 2005). But, in general, France’s export cyclical component is less volatile than Germany’s, which may be associated with the product composition of both countries exports; German exports products have a higher short-term elasticity. Hence, divergences in recent trade performance between France and Germany do not seem to be related to the cyclical part of trade flows. What about the trend in trade flows?

Figure 2 shows annual trend growth of exports. Looking at exports by destination, it seems that Germany has benefited more from the excellent economic performance of China than France. Starting in 2002–03, French export performance is also weak relative to Germany in terms of exports to the EU, the euro area, the United States, and the United Kingdom. France’s export performance is also weaker relative to its own past. In the 1980s, French trend export growth dominated Germany’s only with respect to China; the reverse was true in the 1990s (Table 1). In the 2000s, France’s trend export growth rate, while higher than in the 1990s, was less than 60 percent of Germany’s.

Table 1.

Trend Exports per Destination 1/

(Average annual percent change)

article image

Numbers in bold indicate a higher growth rate of French trend exports.

Table 2.

Trend Exports per Product SITC 1/

(Average annual percent change)

article image

Numbers in bold indicate a higher growth rate of French trend exports.

The deterioration of French export performance vis-à-vis its own past and relative to Germany can be related to products exported. In the 1980s, French trend export growth dominated Germany’s in primary products, chemicals, and miscellaneous manufactured products; the situation was almost the opposite in the 1990s. In the 2000s, of the traditional French exports, France’s trend export growth rate was higher than Germany’s only in chemicals and “other” goods.

The analysis of trend growth rates suggests that there has been since 2002–03 a clear underperformance of French exports relative to the past and also relative to Germany. That France seems less competitive in recent years does not seem to be related to the euro; France’s underperformance is quite broad from a product viewpoint. The change in export performance is relatively recent, but has been protracted enough so as to raise the question of the competitiveness of the French economy. More analysis and time is needed, however, to conclude that there is a structural issue.

IV. Analytical Part: ULCM and TOT Shocks

A. The Model and Economic Conditions for Shocks Identification

To gain further insight into the possible causes of the deteriorating performance of the French foreign sector, this study uses a large dimensional approximate dynamic factor model following the static factor model of Stock and Watson (1998 and 2002).6 The methodology for estimating of the model comprises two main steps. First, estimating the common components of a large panel of data, and second, identifying a reduced number of structural shocks that explain the common components of the variables of interest.7 Once a decision is taken on the process followed by the common components, structural shocks have to be identified. The identification of structural shocks is achieved by focusing on the reduced form VAR residuals. Following Eickmeier (2007), the identification scheme has three steps. First, maximize the variance of the forecast error of the chosen variable and calculate impulse-response functions. Second, the identified shocks are assumed to be linearly correlated to a vector of fundamentals. Finally, orthogonal shocks are identified by rotation using a sign-identification strategy imposing inequality restrictions on the impulse-response functions of variables based on a typical aggregate demand/aggregate supply framework.8 Only those rotations among all possible rotations that have a structural meaning are chosen.

The choice of the variables of interest was motivated by two observations. First, France economic activity is largely influenced by developments in the rest of the world. Thus, it seemed natural to identify a terms of trade (TOT) shock to contrast and compare the behavior of France relative to Germany. Second, in the period of concern, only using unit labor cost measures of the REER, can be seen that French competitiveness deteriorated against Germany in the euro area, although it improved against some other countries. Wages have increased faster in France particularly at the bottom of the scale; these increases have been only partially compensated by higher productivity growth (text figure). Therefore, the second shock that was identified was a shock to unit labor costs in manufacturing (ULCM). The choice seemed also relevant in view of the results of the previous section.

UF2

Real effective exchange rates

(ULC based, 2002=100)

Citation: IMF Working Papers 2009, 002; 10.5089/9781451871494.001.A001

The text table displays the sign restrictions for the identification of shocks that are imposed contemporaneously and during the first year after the shock.

As in major standard macroeconomic models, an increase in ULCM can be interpreted as the result of a fall in labor productivity or an increase in labor compensation. The former is going to be interpreted as a supply shock and the later as a demand shock. This is consistent with the empirical observation that real wages are procyclical. Similarly, a rise in the TOT can result from a deterioration of the country’s competitiveness related to structural factors or alternatively from strong world demand for the country’s products. If the shock is persistent, it will result in an increase in consumption (and investment) and the current account will move into deficit. In contrast, if the TOT increase is due to strong world demand for the small country’s products, given the transient nature of the shock, consumers will largely save the windfall and the current account will move into surplus. Savings will increase.

Identification Inequalities

article image

B. Estimation

The first step of the estimation is the determination of the number of factors. The estimation was done assuming that the series follow an approximate dynamic factor model.9 Using Bai and Ng’s (2002) selection criteria, four factors were retained. Not much can be concluded from the inspection of the factors and their loadings, however, because factors are identified only up to a rotation. Moreover, factors can be a linear combination not only of their contemporaneous values, but also of their lags.

Next, the identification of the structural shocks followed the approach of the structural VAR literature. No identification technology is completely foolproof, however. While the identification technology followed in this paper is flexible enough not to require special restrictions to disentangle common shocks from the contemporaneous transmission of regional or country-specific shocks, it does require additional work, for example, to confirm the nature and source of shocks. In order to properly distinguish a global (common) shock from the transmission within the same period of a country- or regional-specific shock, following Eickmeier (2007), this paper does not restrict the impact effect of the shock. Moreover, after identifying two shocks and giving them an economic interpretation, this study performs the same analysis on a data set containing only French variables. It finds that the impulse-responses of the French-only data set and the broader data set are similar, bringing thus further comfort as to the identification of the source of the shocks.

As it is well know in the literature, only two structural shocks could be identified for each variable of interest. The identification procedure proposed by Uhlig (2003) was applied to the common components of France and Germany’s ULCM and TOT so as to find a reduced number of structural shocks that maximizes the explanation of its forecast error variance over 20 periods.

Sign restrictions on impulse response functions were used to provide economic meaning to the structural shocks. Following Peersman (2005), the angle rotations were applied to the first two principal component shocks taking as pairs a supply shock and a demand shock. The bootstrap was made up of 500 draws.10 The impulse-response functions are calculated for the first five years to display the cyclical pattern associated with the structural shocks. Both the median response and a 90 percent bootstrapped confidence band are estimated.

Two final points on identification are necessary. First, the identification strategy followed in this study, by construction, extracts supply and demand shocks that maximize the explained forecast error variance of the common components of ULCM and TOT. Second, the impulse-response functions from a dataset containing only French variables were similar to those of the full sample, especially the supply shocks.

V. Econometric Results

Results are presented in the form of variance decomposition and impulse-response functions. Tables 3 and 4 show the variance decomposition and the forecast error variance of the common components (henceforth, error variance) of French and German variables explained by the two identified shocks to ULCM. Tables 5 and 6 show the same results for the two shocks to TOT. Figures in Annex II show the impulse-response functions of the French and German shocks to ULCM and TOT and their impact on French and German variables. These shocks suffice to explain up to 99 percent of the error variance of the common components of French and German ULCM over 20 quarters; similarly, these shocks explain up to 97 percent and 99 percent of the error variance of the common component of France and Germany TOT over 20 quarters, respectively. The variance shares of ULCM common components are high as they reach about 75 percent for both countries. In contrast, the variance shares of TOT are much smaller, especially for France: up to 10 percent and 42 percent for France and Germany, respectively. The later suggests that France’s TOT are more influenced than Germany’s by idiosyncratic factors. This is consistent with Kabundi and Nadal De Simone (2007) results: the TOT play a relatively lower role as channels of transmission of international disturbances in France than in Germany.

Table 3.

Forecast Error Variance of the Common Components of French Variables Explained by the Supply and Demand Shocks to Unit Labor Costs in Manufacturing, 1980-2006 1/

article image

Forecast horizon is 20 quarters and refers to the levels of the series. Confidence intervals are constructed using bootstrapping methods.

Table 4.

Forecast Error Variance of the Common Components of German Variables Explained by the Supply and Demand Shocks to Unit Labor Costs in Manufacturing, 1980-2006 1/

article image

Forecast horizon is 20 quarters and refers to the levels of the series. Confidence intervals are constructed using bootstrapping methods.

Table 5.

Forecast Error Variance of the Common Components of French Variables Explained by the Supply and Demand Shocks to Terms of Trade, 1980-2006 1/

article image

Forecast horizon is 20 quarters and refers to the levels of the series. Confidence intervals are constructed using bootstrapping methods.

Table 6.

Forecast Error Variance of the Common Components of German Variables Explained by the Supply and Demand Shocks Terms of Trade, 1980-2006 1/

article image

Forecast horizon is 20 quarters and refers to the levels of the series. Confidence intervals are constructed using bootstrapping methods.

The demand shocks to ULCM and TOT are relatively more important than supply shocks for both countries. Supply and demand shocks have qualitatively broadly similar responses in France and in Germany. However, the quantitative effects as well as the adjustment process are significantly different.

In both countries, supply shocks to ULCM reduce output, private consumption, investment and the volume of exports of goods and services. Employment falls, despite some downward adjustment of real wages. The real exchange rate appreciates. The consumer price index, however, clearly falls in Germany while it is flat in France. The impulse-response functions show that the negative effect on output and on the volume of exports and employment of supply shocks is larger in France than in Germany (there seems to be a relatively larger downward rigidity of wages in France). The SMIC has a tendency to rise somewhat despite the fall in labor productivity. While the dollar value of exports to all destinations increases in Germany, this is not the case in France (e.g., exports to the United States clearly fall). The total increase in the dollar value of French exports is half that of German exports. The same results are evident in terms of the euro value of exports per product, especially for manufactures, transport equipment and mineral fuels and lubricants. France’s euro value of exports is larger than Germany’s for beverages and tobacco, animal and vegetable oils, and commodities and transactions n.e.c. Therefore, France adjusts relatively less via price and wage changes, and more via employment changes than Germany.

Demand shocks to ULCM affect France and Germany differently. A demand shock to ULCM in France produces a short-term small increase in output while employment, real wages and the consumer price index rise without denting productivity. Exports volume tends to increase somewhat while the real exchange rate tends to depreciate. However, as productivity declines, the process is reversed. The value of exports to all destinations and for all products falls. In Germany, the same shock has a much shorter positive impact on output and employment, i.e., less than a year. The consumer price index increases much less than in France; the real wage increase is short lived and gets undone already after 1½ years. Exports volume decrease and the real exchange rate appreciates. The value of exports is not much affected. So, when ULCM increase due to demand pressures, the German economy adjusts more rapidly and seems to display less cost inertia. The real exchange rate helps to offset the negative effects on output and exports while in the case of France it magnifies them.

TOT shocks affect France less than Germany and that difference is more marked following a demand shock than following a supply shock. Positive supply shocks to TOT increase output, investment, and the volume of exports of goods and services. Employment rises, but in France it does so only after real wages have fallen somewhat, given that labor productivity does not change much. In Germany, employment rises sooner and more than in France; the German increase in labor productivity is relatively larger and offsets the rise in real wages enough so that ULCM fall. The real exchange rate depreciation is similar in both economies in the medium run, but it takes longer to reach that level in France than in Germany. The consumer price index falls somewhat in France and is flat in Germany. The dollar value of exports to all destinations has a tendency to fall in France, but the fall is more pronounced in Germany due to the larger short-run exchange rate depreciation experienced by the economy. Exports by product in euros show no major clear patterns, but there is in general a slight increase. Summarizing the results, supply shocks that increase the terms of trade are more consistent with a persistent supply shock in Germany than in France.

Positive demand shocks to TOT result in a negative output effect in France and are clearly inflationary. The real effective exchange rate appreciates as productivity falls and ULCM rise. The SMIC rises despite the fall in labor productivity. The dollar value of French exports by destination increases, except the value of exports to the United States and to accession countries. The increase is, however, larger for Germany, except in terms of exports to China. The euro value of French exports increases less than German exports. In fact, France’s exports are largely flat, except for crude materials, animal and vegetable oils, chemicals and commodities and transactions n.e.c. Overall, the results suggest that the French economy adapts less quickly to inflationary pressures on TOT as a result of a world demand.

A. The 1990s Until Today

The variance shares of ULCM and TOT remain basically the same for France, i.e., around 73 and 10 percent, respectively, during the shorter sample covering the period 1993–2006 (Table 7). The demand shocks to ULCM and TOT are still more significant than supply shocks. The relative importance of the channels of transmission changed. The variance shares of labor productivity and total factor productivity doubled; the variance shares of real compensation of employees, employment, and the SMIC also increased, while the share of consumer prices fell. The results suggest that most variables (except the price level) have become less influenced by idiosyncratic factors. In addition, the error variances indicate that in the recent sample, the role of demand shocks has increased. Similarly, the fall in the variance share of exports from 81 percent to 70 percent suggests that the foreign sector idiosyncratic factors play a more significant role in recent times, a result consistent with the analysis above.

Table 7.

Forecast Error Variance of the Common Components of France Variables Explained by the Supply and Demand Shock to ULCM, 1993-2006 1/

article image

Forecast horizon is 20 quarters and refers to the levels of the series. Confidence intervals are constructed using bootstrapping methods.

Table 8.

Forecast Error Variance of the Common Components of Germany Variables Explained by the Supply and Demand Shock to ULCM, 1993-2006 1/

article image

Forecast horizon is 20 quarters and refers to the levels of the series. Confidence intervals are constructed using bootstrapping methods.

The relative more predominant role played by French idyosincratic factors becomes more obvious when France is compared with Germany. Germany had an increase in the variance share of ULCM and a slight decrease in the variance share of TOT. However, the price level, real compensation of employees, and productivity remained basically unchanged indicating that the role of idyosincratic factors did not change despite reunification. Idyosincratic factors have become more relevant for Germany in terms of the real effective exchange rate, as illustrated by a decline in its variance share. In contrast to France, German exports variance share increased significantly in the more recent period suggesting a more important role for common factors. Like for France, German demand shocks outweigh supply shocks.

VI. Conclusion and Policy Implications

French economic activity is significantly affected by economic activity in the rest of the world. One key channel for the transmission of shocks across countries is international trade. In recent years, the export performance of the French economy relative of its own past and relative to a major trading partner, Germany, has deteriorated. Therefore, the question arises as to whether France is suffering from a competitiveness problem. So far, traditional variables explaining international trade have proved to be insufficient to elucidate the recent decline in France’s export performance. Residuals from econometrically-estimated equations indicate a substantial drag on exports since 2001–02, not attributable to the standard global demand and price/cost factors. In addition, equilibrium exchange rate analysis indicates that France’s real effective exchange rate is largely in line with fundamentals.

This study has found that the recent deterioration of French export performance does not seem to be the related to the “cycle” but to the trend growth of exports, which seems lower in the early 2000s than it was in the past. France’s weaker export performance in the 2000s is reflected both in terms of geographical destination and in terms of product composition.

Given the exposure of the French economy to the rest of the world as well as the known asymmetry in the transmission of disturbances, it was natural to analyze the response of the French economy to typical domestic and foreign shocks. The analysis of the effects of an increase in unit labor costs in manufacturing and of an increase in the terms of trade, suggests that the French economy is relatively less flexible to adjust than the German economy. Faced with an upward shift in unit labor costs, France adjusts relatively less via price and wage changes, and more via employment changes. The same differences are also evident when both countries are faced with an upward TOT shock. To the extent that the convergence of the SMIC operated between 2003 and 2006 represented a significant increase in unit labor costs, and to the extent that the country is a price taker in most of its exports, the study supports the view that the difficulties observed in the French foreign sector may be structural.

The importance of trade flows and relative price changes in the international transmission of disturbances—as well as the policy constraints imposed by the euro area—highlight the relevance of domestic price flexibility. The French economy would benefit from further structural reforms that increase its good, service, and labor markets’ flexibility. This will matter for the magnitude of the real effective exchange rate changes, trade flows, and the size of the current account balance that will be necessary to accommodate the given disturbance.

Similarly, the analysis highlights the importance of measures that increase productivity and, in particular, the desirability of avoiding SMIC adjustments unrelated to productivity.

Appendix I. The Corbae-Ouliaris Ideal Band-Pass Filter

Let us assume that Xt is an I(1) process with ΔXt = vt such that vt has a Wold representation. The spectral density of vt is fvv (λ) >0, for all λ. The discrete Fourier transform of Xt for λt ≠ 0:

wX(λs)=11−eiλswv(λs)−eiλs1−eiλs(Xn−X0)n1/2,

where λs=2πsn, s = 0, 1, …, n-1, are the fundamental frequencies. The second term makes it clear that the Fourier transform is not asymptotically independent across fundamental frequencies because the second term is a deterministic trend in the frequency domain with a random coefficient (Xn−X0)n1/2 Unless that term is removed, it will produce leakages into all frequencies λt ≠ 0, even in the limit as n →∞. Sacrificing a single observation, instead of estimating the random coefficient a la Hannan (1970), Corbae and Ouliaris (2006) show that by imposing that (Xn − X1) = (Xn − X0) will produce an estimate that will have no finite sampling error, has superior endpoint properties, and has much lower mean-squared error than popular time-domain filters such as HP or B-K. In addition, in contrast to B-K, it is consistent. This is the ideal band-pass filter used in the paper.

Appendix II. The Approximate Dynamic Factor Model

This study uses a large dimensional approximate dynamic factor model in the tradition of Stock and Watson (1998 and 2002). In contrast to the models of Sargent and Sims (1977) and Geweke (1977), it admits the possibility of serial correlation and weakly cross-sectional correlation of idiosyncratic components, as in Chamberlain (1983) and Chamberlain and Rothschild (1983). Similar models have recently been used by Giannone, Reichlin, and Sala (2002), Forni and others (2005), and Eickmeier (2006 and 2007).

A vector of time series Yt=(y1t, y2t,…, yNt)′ can be represented as the sum of two latent components, a common component Xt =(x1t,x2t,…,xNt)′ and an idiosyncratic component Ξt =(ε1t, ε2t,…, εNt)‘

Yt=Xt+Ξt(1)Yt=CFt+Ξt

where Ft = (f1t, f2t,…, frt)′ is a vector of r common factors, and C=(c1′,c2′,…,cN′)′ is a N × r matrix of factor loadings, with r < <N. The common component Xt, which is a linear combination of common factors, is driven by few common shocks, which are the same for all variables. Nevertheless, the effects of common shocks differ from one variable to another due to different factor loadings. The idiosyncratic component is driven by idiosyncratic shocks, specific to each variable. The static factor model used here differs from the dynamic factor model in that it treats lagged or dynamic factors Ft as additional static factors. Thus, common factors include both lagged and contemporaneous factors.

Using the law of large number (as T, N → ∞), the idiosyncratic component, which is weakly correlated by construction, vanishes; and therefore, the common component can be easily estimated in a consistent manner by using standard principal component analysis. The first r eigenvalues and eigenvectors are calculated from the variance-covariance matrix cov(Yt).

Xt=VV′Yt,(2)

and since the factor loadings C =V, equation (1) becomes,

Ft=V′Yt.(3)

From (1), the idiosyncratic component is

Ξt=Yt−Xt.(4)

From all the more or less formal criteria to determine the number of static factors r, Bai and Ng (2002) information criteria was followed. As in Forni and others (2005), Ft was approximated by an autoregressive representation of order 111:

Ft=BFt−1+ut,(5)

where B is a r × r matrix and ut a r × t vector of residuals. Equation (5) is the reduced form model of (1).

Once a decision is taken on the process followed by the common components, structural shocks have to be identified by focusing on the reduced form VAR residuals of (5). Following Eickmeier (2007), the identification scheme has three steps. First, maximize the variance of the forecast error of the chosen variable and calculate impulse-response functions. The interest here is unit labor costs in manufacturing (ULCM) and terms of trade. So, using ULCM as an example, a few major shocks driving unit them are identified.12 This implies maximizing the explanation of the chosen variance of the k-step ahead forecast error of ULCM with a reduced number of shocks.13 To this end, k -ahead prediction errors ut are decomposed into k mutually orthogonal innovations using the Cholesky decomposition. The lower triangular Cholesky matrix A is such that ut = Avt and E(vtvt′)=I Hence,

cov(ut)=AE(vtvt′)A′=AA′.(6)

The impulse-response function of yit to the identified shock in period k is obtained as follows:

Rik=ciBkA,(7)

with ci the ith row of factor loadings of C and with a corresponding variance-covariance matrix ∑j=0kRijRij′.

Second, the identified shocks are assumed to be linearly correlated to a vector of fundamentals. The fundamental forces ωt =(ω1t, ω2t,…, ωrt)′ behind France’s ULCM are correlated to the identified shocks through the r × r matrix Q. Thus,

vt=Qωt.(8)

The intuition of the procedure is to select Q in such a way that the first shock explains as much as possible of the forecast error variance of the France’s ULCM common component over a certain horizon k, and the second shock explains as much as possible of the remaining forecast error variance. Focusing on the first shock, the task is to explain as much as possible of its error variance

σ2(k)=∑j=0k(Rijq1)(Rijq1)′,(9)

where i is, in our example, the French ULCM, and q1 is the first column of Q. The column q1 is selected in such a way that q1′σ2q1 is maximized, that is

σ2(k)=∑j=0k(Rijq1)(Rijq1)′=q1′Sikq1

where Sik=∑j=0k(k+1−j)Rij′Rij.

The maximization problem subject to the side constraint q1′q1=1, can be written as the

Lagrangean,

L=q1′Sikq1−λ(q1′q1−1),(10)

where λ is the Lagrangean multiplier. From (10), q1 is the first eigenvector of Sik with eigenvalue λ and, therefore, the shock associated with q1 is the first principal component shock. Q is the matrix of eigenvectors of S, (q1, q2, …, qr), where ql (l=1,…,r) is the eigenvector corresponding to the lth principal component shock. Along the lines of Uhlig (2003), Eickmeier (2007), and Altig and others (2002), it is posed: k = 0 to k = 19, i.e., five years, which covers short- as well as medium-run dynamics.

Finally, orthogonal shocks are identified by rotation. If two shocks are identified, following Canova and de Nicoló (2003), the orthogonal shocks vector ωt = (ω1t,ω2t)′ is multiplied by a 2 × 2 orthogonal rotation matrix P of the form:

P=(cos(θ)−sin(θ)sin(θ)cos(θ)),

where θ is the rotation angle; θ ∊ (0,π), produces all possible rotations and varies on a grid. If θ is fixed, and q = 5, there are q(q − 1)/2 bivariate rotations of different elements of the VAR. Following the insights of Sims and Zha (1999), and as in Peersman (2005), Canova and de Nicoló (2003), Eickmeier (2007), Kabundi and Nadal De Simone (2007) the number of angles between 0 and π is assumed to be 12: this implies 6,191,736,421x1010 (1210) rotations. Hence, the rotated factor wt = Pwt explains in total all the variation measured by the first two eigenvalues. This way the two principal components ωi are associated to the two structural shocks wi through the matrix P, and the impulse-response functions of the two structural shocks on all the fundamental forces can be estimated.

A sign-identification strategy is followed to identify the shocks. The method was developed by Peersman (2005). This strategy imposes inequality sign restrictions on the impulse response functions of variables based on a typical aggregate demand and aggregate supply framework.14 Only those rotations among all possible q × q rotations that have a structural meaning are chosen. The text table displays the sign restrictions for the identification of shocks that are imposed contemporaneously and during the first year after the shock.

Identification Inequalities

article image