France: Selected Issues
Author: International Monetary Fund
Publication Date:
07 Nov 2005
eISBN:
9781451813654
Language:
English
Keywords:
ISCR; CR; labor market; worker; core inflation; inflation profile; underlying inflation behavior; inflation pressure; rises end; inflation component; Transportation; Oil prices; Housing; Real estate prices
Download PDF (382.8 KB)

This Selected Issues paper analyzes the recent French inflation behavior. The paper demonstrates that the recent change in French headline inflation behavior resulted from a few idiosyncratic, i.e., transient, factors. The paper describes the model setup and calibration of some crucial economic relations and parameters. The level of competition in labor and product markets in France and the other EU countries is discussed, including the size of markups. The paper also looks into the economic impact of increasing competition in each market separately and the advantages of coordinating reforms within the euro area.

Abstract

This Selected Issues paper analyzes the recent French inflation behavior. The paper demonstrates that the recent change in French headline inflation behavior resulted from a few idiosyncratic, i.e., transient, factors. The paper describes the model setup and calibration of some crucial economic relations and parameters. The level of competition in labor and product markets in France and the other EU countries is discussed, including the size of markups. The paper also looks into the economic impact of increasing competition in each market separately and the advantages of coordinating reforms within the euro area.

I. Recent French Inflation Behavior: Is it Any Different from the Euro Area’s?1

A. Introduction

1. French headline inflation moved above the euro area average during the last two years. Measured as year-on-year changes in the monthly Harmonized Index of Consumer Prices (HICP), French inflation, traditionally below the euro area average, moved above it around the beginning of 2003 and remained so until the start of 2005, when it fell back below the euro area average (Figure 1). The change in French inflation behavior becomes even clearer using a measure of core inflation—the HICP excluding energy and unprocessed food. In contrast, a similar change in inflation behavior is not apparent in the three other major euro area countries, Germany, Italy, and Spain.

Figure 1.
Figure 1.

France: Headline and Core Inflation

(Percent change over same period of previous year)

Citation: IMF Staff Country Reports 2005, 397; 10.5089/9781451813654.002.A001

Source: Cronos database.

2. This chapter demonstrates that the recent change in French headline inflation behavior resulted from a few idiosyncratic, i.e., transient factors. Applying the generalized dynamic factor model (GDFM) to a disaggregated panel data set, it appears that beyond the impact of oil prices on headline inflation, the change in inflation behavior is related to two sets of factors. The first one is government policies, as reflected in the role of tobacco excise taxes, electricity, and hospital service prices. The second factor is what could be referred to as specific market conditions and developments. Given the recent buoyancy of the French real estate market and residential construction, it is not surprising that housing rents come second on the list of idiosyncratic components of inflation. In addition, with the adoption of the 35-hour workweek, wage price pressures can be expected to be the strongest in sectors with pricing power. This could explain the higher-than-average inflation in domestic services, maintenance and repair of personal transport equipment and dwellings. Despite some cross-country variation, results for the other three major euro area countries also point to the role of oil prices and indirect taxes, but not to the extent of changing the inflation profile as in France.

3. Contingent on idiosyncratic inflationary factors subsiding, French headline inflation can be expected to decelerate. There are already signs of a deceleration of house price inflation and a more moderate pace of activity in the construction sector, which should slow down the progression of rental housing costs. However, a durable decline of headline inflation will require that oil price rises end and that increases of indirect taxes and public sector-controlled prices moderate. The latter points to the possibility of enhancing welfare by better coordinating policy objectives such as sustaining private demand with correcting fiscal imbalances, which may require tariff or excise tax hikes.

4. As a corollary, this chapter argues that the common components of inflation are a more robust indicator of underlying inflation pressures than the standard core inflation measure. The common components of inflation are less clouded by large idiosyncratic factors than the standard measure of core inflation. As such, they may be superior both for formulating and providing accountability in monetary policy-making.

5. The remainder of the chapter is organized as follows. Section B presents the data and the model to be estimated. Section C discusses the main results regarding the common and idiosyncratic components. Section D proposes the common components of headline inflation as a robust measure of underlying inflationary pressures. Section E concludes.

B. The Data and the Model

6. This study uses HICP data for 12 member countries of the euro area and the euro area aggregate. The HICP data are at the most disaggregated level currently available, i.e., the 4-digit level, for 12 members of the euro area (11 members only until December 2000). The data panel comprises 1,177 series with 99 observations. The sample period is 1996M1–2005M3, the longest period for which the HICP is available for at least 11 euro area member countries. Inflation is defined as annual inflation measured as the change in the HICP between a given month and the same month of the previous year.2

7. Distinguishing between underlying and transitory inflationary forces is often fraught with imprecision and practical difficulties. Traditionally, monthly inflation reports take a bottom-up (i.e., disaggregated) approach to “explain” inflation and signal out the main “culprits.” They refer, for instance, to the role of exogenous factors such as an unusually early end of sales, a rapid increase in oil prices, or a policy-driven factor such as a rise in tobacco excise taxes. Yet, a proper distinction between underlying and transient inflationary forces is not always clear. For instance, statistical approaches such as a trimmed-mean involve a great deal of subjectivity. Also, while using the median may take care of skewness in price changes, it does not necessarily take into account that price changes tend to be kurtotic, an issue discussed further in Section D.3 A more structured procedure is desirable, especially one that is flexible enough to be used regularly for policy analysis and forecasting at a minimum computational cost.

8. This chapter uses the GDFM to distinguish persistent underlying inflationary forces from transient idiosyncratic ones, and discusses the economic developments that may explain the later. Forni and others (2000 and 2003) expanded on the principal component and the Stock and Watson’s (1989) methods by developing a coincident and a leading indicator called the GDFM. The GDFM reconciles dynamic principal components analysis with the dynamic factor model of Sargent and Sims (1977) and Geweke (1977). This method allows an estimation of an index without an a priori distinction between coincident and noncoincident series. Unlike Stock and Watson’s approach (1989), Forni and others (2000 and 2003) point out that leading and lagging variables contribute to a better estimation of the coincident indicator. The coincident index is the weighted average of common components.4 The GDFM differs from other methods in that it allows for limited cross-correlation among idiosyncratic components; imposing zero cross-correlation is a demanding restriction that seems unrealistic in most practical cases. Identification is achieved by working with a large data panel.5

9. The GDFM enables the efficient estimation of the common and idiosyncratic components of very large data sets. The GDFM assumes that each time series in a large data set is composed of two sets of unobserved components. First, the common components, which are driven by a small number of shocks that are common to the entire panel—each time series has its own loading associated with the shocks. Second, the idiosyncratic components, which are specific to a particular variable and orthogonal with the past, present, and future values of the common shocks. Common components of inflation are best viewed as the result of the underlying inflation process, and it is thus expected that they will be persistent.6 The idiosyncratic components instead reflect local aspects of markets that are not persistent, i.e., those that affect a specific industry such as an increase in the price of fresh fruit following a drought. The idiosyncratic components, however, also capture unsystematic measurement errors. Therefore, even though idiosyncratic components are of limited interest for medium- and long-run movements of inflation, their role is far from negligible and they may account for a sizable percentage of inflation in the short term. If idiosyncratic components reflect truly local aspects of markets, they will display little correlation with the rest of the variables in the panel. However, note that to the extent that idiosyncratic components are allowed to be mutually correlated to some extent in the model, substitution-on-demand effects, for instance, will result in some “contagion” across idiosyncratic components.

10. The GDFM model is very general. Assume a vector of n series expressed as in equation (1):

xti=αi(L)ut+ɛti,(1)

where xti is an (t×n) vector stochastic stationary process with zero mean and variance 1, xti=(xtli,xt2i,…,xtni)′;ut1,ut2,…,utq)′ is a (q×t) vector of mutually orthogonal common shocks with zero mean and unit variance, and with q<n;ɛti,=(ɛt1i,ɛt2i,…,ɛtni)′ is a (n×t) vector of idiosyncratic shocks; and αi (L)' is a (n×q) matrix of rational functions with the lag operator L. The model allows for correlation between ɛti variables, but the variances of ɛti are bounded as i→∞. When n is large, the idiosyncratic components, which are poorly correlated, will vanish, and only the common components will be left, and thus they will be identified (see Forni and others, 2000, for a technical proof).

11. The GDFM model is estimated using the one-sided estimator proposed by Forni and others (2003). The procedure comprises two steps:7 first, estimating the spectral density matrix of the vector stochastic process xti and, second, using the calculated q largest (real) eigenvalues—and their corresponding eigenvectors—of the spectral density matrix to estimate the generalized common components. In this study, the xti (t× n) vector stochastic stationary process has t = 99 monthly observations and n = 1,177 series; q = 1 common shock—underlying inflation. Accordingly, there are 1,177 idiosyncratic shocks; and in the αi (L)’ (n×q) matrix of rational functions with the lag operator L, the number of lags M is 6.8

C. Inflation Common and Idiosyncratic Components

12. The euro area as well as all countries in the sample experienced an increase in headline inflation toward the end of 1999. The increase in the mean of headline inflation of the euro area and all countries toward the end of 1999 is even clearer when looking at the common components of inflation (Figure 2).

Figure 2.
Figure 2.

France: Headline Inflation Components

(Percent change over same period of previous year)

Citation: IMF Staff Country Reports 2005, 397; 10.5089/9781451813654.002.A001

Sources: Cronos database; and IMF staff calculations.

13. Starting in 2003, French headline inflation moved above the euro area average, but this was not the result of a persistent shock to inflation. No country in the sample experienced any change in the profile of its inflation common components relative to the euro area during the sample period. The common components of French and German inflation have been consistently below the common components of euro area inflation. In contrast, Italian and Spanish inflation common components have consistently been above the euro area inflation common components. With common components reflecting underlying inflation behavior, the change in French inflation behavior that started at the beginning of 2003 did not signal a persistent shock to inflation.

14. The recent increase in French headline inflation was the result of idiosyncratic components. French idiosyncratic factors added an average of about ⅔ of one percentage point to annual inflation starting in 2003.9 Instead, German idiosyncratic components of inflation became negative in the second half of 2002 and remained so until the first half of 2004, subtracting about ⅓ of one percentage point from headline inflation during that period. Also starting in 2003, Italian idiosyncratic inflation components added about ¼ of one percentage point to headline inflation per annum, while Spanish idiosyncratic inflation components added just 1/10 of one percentage point to inflation.

15. A few domestic factors explain the recent increase in headline inflation in France (Figure 3 and Table 1). Tobacco is the item that recorded the highest inflation rate by far; it is followed by housing rents, electricity, spare parts and accessories for personal transport equipment, and a number of service-related items. Oil prices seem to have a particularly strong effect on French inflation.10 Given their nature, however, idiosyncratic components or local factor inflation should vanish, even in the medium run, which could be taken to be the sample period used in this study. Indeed, the 1997M1–2005M3 average idiosyncratic inflation in each of the items is basically zero (not shown). Over a relatively shorter period of time, idiosyncratic factors will not be zero, however; structural factors such as the demand price elasticity, pricing power, and government policy are therefore bound to play a major role in explaining the idiosyncratic components’ behavior and thus affect headline inflation. In what follows, the main possible factors explaining the idiosyncratic components of inflation are discussed.

Figure 3.
Figure 3.

France vs. Euro Area Selected Idiosyncratic Components of Headline Inflation

(Percent change over same period of previous year)

Citation: IMF Staff Country Reports 2005, 397; 10.5089/9781451813654.002.A001

Sources: Cronos database; and IMF staff calculations.
Table 1.

France and Euro Area: Idiosyncratic Components of Inflation 1/

article image
Sources: Cronos database; and IMF staff calculations.

The table displays the items which idiosyncratic components recorded the highest inflation since 2003M1 relative to the euro area idiosyncratic components. Only those HICP items that recorded idiosyncratic inflation rates larger than the average idiosyncratic inflation are included.

16. Government policies is one of the two main domestic factors explaining the idiosyncratic components of French inflation. Government policies affect tobacco prices, which have been subject to large excise-tax-related price increases during the last two years. The monopoly taxing power of the government is clearly evident in the idiosyncratic component, reflecting as well the low short-run price elasticity of demand for tobacco. Similarly, the role of the French state in price formation seems to explain idiosyncratic components of inflation in the electricity sector, in hospital services, and in medical and paramedical services. The privatization of telephone and telefax services undertaken in France had a positive impact on inflation as the corresponding idiosyncratic component shows.

17. Specific market conditions are the other domestic factor explaining idiosyncratic components of French inflation. Specific market conditions include two groups of items:

  • Wage pressures in sectors that employ mostly unskilled workers may explain part of French idiosyncratic inflation. The impact of the 35-hour workweek on firms’ operating costs and margins may explain the price increases in sectors with pricing power. This is the case of domestic services and household services, maintenance and repair of personal transport equipment and the dwelling, and refuse collection (Figure 4).

  • Similarly, the buoyant housing market and construction sector contribute to inflationary pressures in France (housing rents come second on the list of idiosyncratic components inflation and are legally tied to the construction price index).

Figure 4.
Figure 4.

France: Employees Earning Minimum Wage and Sectoral Wage Growth

Citation: IMF Staff Country Reports 2005, 397; 10.5089/9781451813654.002.A001

Sources: ACEMO; and Bank of France.

18. Most likely due to cyclical factors, German idiosyncratic components of inflation made a negative contribution to headline inflation (Table 2). Despite the increase in tobacco excise taxes in 2003, the rise in gas and electricity prices as well as the price increases that resulted from health care reform, cyclical forces held idiosyncratic components of inflation down. Consequently, for instance, prices of services such as package holidays and restaurants fell. Rents for housing contributed negatively, reflecting the subdued performance of the German real estate market.

Table 2.

Germany and Euro Area: Idiosyncratic Components of Inflation 1/

article image
Sources: Cronos database; and IMF staff calculations.

The table displays the items which idiosyncratic components recorded the lowest inflation since 2003M1 relative to the euro area idiosyncratic components. Only those HICP items that recorded idiosyncratic inflation rates lower than the average idiosyncratic inflation are included.

19. Public policy-related price increases are key idiosyncratic components of recent Italian inflation (Table 3). As in France, excise tax increases explain the large contribution of tobacco to Italian idiosyncratic components of inflation.11 In addition, both the financial difficulties of the Italian national carrier and the contractual changes that took place between air companies and travel agencies’ commercial relationships resulted in an increase in air fares. Additional taxation of air travel in 2004 also contributed. Local governments hiked prices for water supply and local transport. Finally, new financial sector regulations increased costs, which were reflected in higher prices for financial services.

Table 3.

Italy and Euro Area: Idiosyncratic Components of Inflation 1/

article image
Sources: Cronos database; and IMF staff calculations.

The table displays the items which idiosyncratic components recorded the highest inflation since 2003M1 relative to the euro area idiosyncratic components. Only those HICP items that recorded idiosyncratic inflation rates larger than the average idiosyncratic inflation are included.

20. Various factors, including government-related price increases, contributed to Spanish idiosyncratic inflation, which was, however, quite small (Table 4). A drought is largely responsible for foodstuff price inflation (e.g., fruit and cereals), and government-related price increases pushed inflation (e.g., electricity). Yet, as discussed below, there may be some market idiosyncrasies associated with the catching-up process of prices to the euro area average in areas such as services (e.g., restaurants).

Table 4.

Spain and Euro Area: Idiosyncratic Components of Inflation 1/

article image
Sources: Cronos database; and IMF staff calculations.

The table displays the items which idiosyncratic components recorded the highest inflation since 2003M1 relative to the euro area idiosyncratic components. Only those HICP items that recorded idiosyncratic inflation rates larger than the average idiosyncratic inflation are included.

D. Corollary: Common Components of Inflation and Underlying Inflation

21. The estimated model identifies the common components of inflation or underlying inflationary forces, and so does core inflation: which is a better measure of trend inflation? As in the euro area, the common components of French inflation suggest that underlying inflation has been increasing since end-December 2004. The core inflation measure, instead, points to a decline (Figures 1 and 2). Which indicator is superior as a measure of trend, underlying inflation? The question is best viewed from the perspective of statistical inference as price changes are not drawn from a well-known, stable population distribution. If price changes were approximately normal, then the sample mean would be the best estimator of trend inflation in the sense of being unbiased and efficient. If price changes are not normal, however, the mean may still be an unbiased estimator of underlying inflation, but it may not be as robust or efficient.

22. Common components have different statistical properties from core inflation. Various statistical measures illustrate that common components and core inflation (normally referred to as “underlying inflation” and calculated as headline inflation excluding energy and unprocessed food) have different statistical behavior (Figure 5). The policy horizon that interests the analyst or monetary policymaker matters for deciding whether the common components of inflation are a more robust measure of underlying inflation than the traditional core inflation measure, both for formulating policy and for policy accountability.

Figure 5.
Figure 5.

France: Spectral Analysis of Headline and Core Inflation and Headline Common Components

(In percent)

Citation: IMF Staff Country Reports 2005, 397; 10.5089/9781451813654.002.A001

Sources: Cronos database; and IMF staff calculations.

23. If a monetary policymaker is interested either in trend inflationary pressures or in inflation at periodicities of one to two years, the common components of inflation may be a robust measure of underlying inflation.12 First, headline inflation comoves more closely with its common components than with the standard measure of core inflation; central banks are interested in a measure of underlying inflation that minimizes the influence of all supply shocks so as to achieve their inflation objectives with minimal output variance.13 The comparison with respect to headline inflation is pertinent as headline inflation is the preferred measure of inflation by central banks that make explicit their inflation targets, and similarly, the yardstick by which they are made accountable. Second, core inflation simply assumes that the two items excluded are uncorrelated with underlying inflation and also that they are the only items that mostly respond to supply shocks. Instead, the common components are constructed imposing the constraint that idiosyncratic components are uncorrelated with them, and this assumption is applied to all items in the HICP. Because common components provide a clearer measure of underlying inflation, they avoid the impression that a deviation from the target is a weakening in the monetary authority’s commitment to the target. This should also minimize the risk of the supply shock feeding through into inflation expectations.14

24. Future research could explore the relative efficiency of the common components measure given that alternative measures of underlying inflation, such as the trimmed mean or the median, have drawbacks. It is known that the relative efficiency of estimators of trend inflation is sensitive to the kurtosis (i.e., the presence of extreme price changes) of the distribution of price changes. In the case of France, the kurtosis of price changes oscillates between about 5 and 59. Given such high values (the normal distribution has a kurtosis of 3), measures that attribute little (or zero) weight to the tails of the distribution of price changes are bound to be more efficient estimators of trend inflation. One such measure is the trimmed-mean. Simulation studies suggest, however, that the trimmed-mean performs well when kurtosis is between 4 and 5.5 (see Roger, 1997). For higher values of kurtosis, the median is instead recommended. The median, however, is not an efficient estimator of the trend when the population is skewed, which is the case for price changes (e.g., Bryan and Cecchetti, 1996). In France, skewness oscillates between -2½ and 7¼ during the sample period. Future research could explore the relative efficiency of the common components of inflation as estimators of trend, underlying inflation.15

E. Concluding Remarks

25. French inflation behavior changed between the start of 2003 and the beginning of 2005 due to oil price rises, increases in indirect taxation, and specific market developments. Using statistical methods, it was found that there has not been any major change in underlying inflation pressures for France or any other country in the sample. The change in the French inflation profile was thus solely an idiosyncratic phenomenon. It was due to the peculiar impact that increases in oil prices had in France, and to factors associated with indirect taxation and specific market developments such as the booming French real estate market and construction sector. While it is true that shocks such as oil price rises also affected other countries in the sample and the euro area as a whole, the relatively more dynamic French domestic demand may have facilitated the transmission of the shocks to headline inflation.

26. The results in this study suggest that French inflation should decelerate in the near future if the idiosyncratic components of headline inflation subside, while raising the issue of the timing of public sector-induced price changes. Barring further oil prices or government-related price hikes, French headline inflation should return to its underlying trend in the near future. These developments, however, raise the policy question of the timing of tax changes. Increases in indirect taxes or in utilities tariffs, to the extent that they boost inflation, may conflict with other policy objectives simultaneously sought, such as supporting real disposable income and increasing private consumption.

27. As a corollary, this study shows that the common components of French inflation are a more robust indicator of underlying inflation than the standard core inflation measure. Thus, they may be better for formulating and providing accountability in monetary policy-making. They are less clouded by large idiosyncratic factors that could easily creep into the standard measure of core inflation. Moreover, the relative efficiency of the common components of inflation vis-à-vis alternative measures such as the trimmed-mean and the median, is an issue that deserves to be explored.

References

  • Bryan, M. and S. Cecchetti, 1996, “Inflation and the Distribution of Price Changes,” National Bureau of Economic Research, Working Paper no. 4303.

    • Search Google Scholar
    • Export Citation

  • De Bandt, O., A. Saint-Guilhem, H. Le Bihan, and J. Matheron, 2004, “La Persistence de l’Inflation en France et dans la Zone Euro,” Bulletin de la Banque de France, no. 127, pp. 31–50, Banque de France.

    • Search Google Scholar
    • Export Citation

  • Elliott, G., T. J. Rothenberg, and J. Stock, 1996. “Efficient Tests for an Autoregressive Unit Root,” Econometrica 64, pp. 813–36.

    • Search Google Scholar
    • Export Citation

  • Forni, M., M. Hallin, M. Lippi, and L. Reichlin, 2000, “The Generalized Factor Model: Identification and Estimation,” Review of Economic and Statistics, Vol. 82, no. 4, pp. 540–54.

    • Search Google Scholar
    • Export Citation

  • Forni, M., M. Hallin, M. Lippi, and L. Reichlin, 2003, “The Generalized Factor Model: One- Sided Estimation and Forecasting,” manuscript.

    • Search Google Scholar
    • Export Citation

  • Friedman, M., 1963, Inflation. Causes and Consequences, Asia Publishing House, New York.

  • Geweke, J., 1977, “The Dynamic Factor Analysis of Economic Time Series,” in Latent Variables in Socio-Economic Models; edited by D.J. Aigner and A.S. Golberger (Amsterdam: North-Holland), p. 19.

    • Search Google Scholar
    • Export Citation

  • INSEE, 1990-2002, “Enquête Emploi,” Institut National de la Statistique et des Etudes Economiques.

  • Keynes, J. M., 1930, A Treatise on Money, Macmillan, London.

  • Quah, D. and S. Vahey, 1995, “Measuring Core Inflation,” The Economic Journal, Vol. 105, 86 (1), pp. 1130–44.

  • Roger, S., 1997, “A Robust Measure of Core Inflation in New Zealand, 1949-96,” Discussion Paper G97/7, Reserve Bank of New Zealand.

  • Sargent, T.J., and C.A. Sims, 1977, “Business Cycle Modeling without Pretending to have too Much a Priori Economic Theory,” in New Methods in Business Research, edited by C.A. Sims (Mineapolis: Federal Reserve Bank of Mineapolis).

    • Search Google Scholar
    • Export Citation

  • Stock, J.H., and M.H. Watson, 1989, “New Indexes of Coincident and Leading Economic Indicators,” NBER Macroeconomic Annual Report 1989, pp. 351–94.

    • Search Google Scholar
    • Export Citation

  • Yang, X., 2003, “The Role of World Common Shocks in International Economic Fluctuations,” manuscript.

1

Prepared by Francisco Nadal De Simone.

2

The main results are not altered when inflation is defined as the monthly change in the HICP.

3

Kurtosis characterizes the relative peakedness or flatness of a probability distribution compared with the normal distribution.

4

As in Stock and Watson, the common shocks and their factor loadings together make the common components.

5

Technically, to achieve identification, Forni and others (2000 and 2003) assume an infinite cross-sectional data dimension. The idiosyncratic components are stationary processes uncorrelated with past, present, and future values of the common shocks. In addition, the common component of each series can be uniquely identified and consistently estimated if (1) the spectral density matrix of the observation matrix Xt exists; (2) the first q eigenvalues go to infinity as the number of series n goes to infinity; and (3) all remaining eigenvalues are bounded. In practice, when the number of series is larger than 1,000, as in this study, those assumptions may reasonably be expected to hold.

6

Friedman (1963) emphasized the distinction between the steady, trend, or persistent element of inflation, which tends to be incorporated into expectations, and the transient component, which is less readily anticipated. He associated underlying inflation with the persistent component of inflation. Quah and Vahey (1995), however, include also the cyclical part of inflation into their definition of underlying inflation. A subtle different view on underlying inflation may result from associating it with generalized inflation (Keynes, 1930). In this view, relative price changes, while temporary, are expected to also affect aggregate inflation. Only if one believes that relative price changes that affect aggregate inflation are mostly driven by supply shocks, both views of underlying inflation are basically equivalent. This matter can only be settled empirically and is not explored here.

7

The headline inflation series were tested for stationarity using the unit root test developed by Elliott, Rothenberg, and Stock (1996). The tests are available upon request.

8

Forni and others (2000) suggest to determine the number of lags using the formula. However, this formula tends to result in too few lags. For example, in this study, the formula would suggest three lags. Given the high frequency of the data, and Yang’s (2003) suggestion to include more lags, M = 6 in this study so as to capture higher order dynamics.

9

According to the Banque de France, the annual impact of the euro nominal appreciation on French headline inflation was -1.2 percentage points in 2003, the period during which the idiosyncratic components increased French inflation by 0.3 of one percentage point (De Bandt and others, 2004).

10

The difference between the common components of inflation in fuels and lubricants and in liquid fuels between the euro area and France is zero. However, France seems to have a much higher incidence of transitory inflationary forces in those categories than the euro area. The causes for this behavior remain to be established, but could be linked to changes in the excise tax regime weighing on these items in France as well as the relatively more dynamic French domestic demand, which may have facilitated the transmission of the shocks to inflation.

11

The causes for the rapid increase in the idiosyncratic components of restaurant services are not known.

12

A key practical requirement for a measure of underlying inflation is also its timeliness. The common components can be made available almost simultaneously with the publication of the price index, and in this sense are not at a disadvantage with respect to the standard measure of core inflation.

13

The correlation between the common components and HICP is about 70 percent, while the correlation between core inflation and the HICP is 61 percent; the relative similarity of the spectral shapes of headline inflation and common components inflation beyond periodicities of 21 months supports this point.

14

Conceivably, the common components can be useful in forward-looking accountability as well when a shock blurs developments in underlying inflation. For example, if an adverse supply shock increases inflation and masks a fall in underlying inflation, an easing may be misinterpreted as a weakening in the monetary authority’s commitment to its inflation target.

15

While the estimated common components presumably take care of the reweighing of the HICP items associated with the trimmed-mean, it is far from clear that the method does well in accounting for possibly time-varying trends in relative price changes. It is known, in contrast, that with time-varying trends in relative price changes, the trimmed-mean may exclude not only temporary shocks but also part of trend inflation.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
France: Selected Issues
Author: International Monetary Fund
Volume/Issue:
Volume 2005: Issue 397
Publisher:
International Monetary Fund
ISBN:
9781451813654
ISSN:
1934-7685
Pages:
70
DOI:
https://doi.org/10.5089/9781451813654.002
  • View in gallery

    France: Headline and Core Inflation

    (Percent change over same period of previous year)

  • View in gallery

    France: Headline Inflation Components

    (Percent change over same period of previous year)

  • View in gallery

    France vs. Euro Area Selected Idiosyncratic Components of Headline Inflation

    (Percent change over same period of previous year)

  • View in gallery

    France: Employees Earning Minimum Wage and Sectoral Wage Growth

  • View in gallery

    France: Spectral Analysis of Headline and Core Inflation and Headline Common Components

    (In percent)

France: Headline and Core Inflation

(Percent change over same period of previous year)

France: Headline Inflation Components

(Percent change over same period of previous year)

France vs. Euro Area Selected Idiosyncratic Components of Headline Inflation

(Percent change over same period of previous year)

France: Employees Earning Minimum Wage and Sectoral Wage Growth

France: Spectral Analysis of Headline and Core Inflation and Headline Common Components

(In percent)