Greece
Background Paper
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar Close
Publication Date:
03 Oct 1995
eISBN:
9781451816051
Language:
English
Keywords:
ISCR; CR; economic activity; foreign currency; exchange rate; market share; rate of depreciation; marginal utility; trading partner; consumer price index; nominal interest rate
Download PDF (8.5 MB)

This Background Paper reviews economic developments in the real economy and the labor market of Greece during 1990–95. It presents the calculation of an index of coincident economic indicators, which is used to gauge the strength of the recovery. The paper highlights that in 1994, Greece’s external accounts strengthened considerably: the current account deficit improved to near-balance reflecting strong invisible receipts (mostly tourism), and substantial capital inflows—following the successful handling of the speculative attack against the drachma in May 1994—boosted official reserves to a record US$16.5 billion at end-June 1995.

Abstract

This Background Paper reviews economic developments in the real economy and the labor market of Greece during 1990–95. It presents the calculation of an index of coincident economic indicators, which is used to gauge the strength of the recovery. The paper highlights that in 1994, Greece’s external accounts strengthened considerably: the current account deficit improved to near-balance reflecting strong invisible receipts (mostly tourism), and substantial capital inflows—following the successful handling of the speculative attack against the drachma in May 1994—boosted official reserves to a record US$16.5 billion at end-June 1995.

I. Introduction

In 1994 and 1995, economic activity in Greece started to turn around: output is projected to grow by almost 2 percent this year, and unemployment to stabilize under 10 percent. Chapter II (prepared by A. Hamann) reviews recent developments in the real economy and the labor market, and Chapter III (prepared by C. Christofides) presents the calculation of an index of coincident economic indicators, which is used to gauge the strength of the recovery.

Nevertheless, after a marked deterioration in the 1970s and 1980s, Greece’s long-term growth performance continues to be disappointing. In Chapter IV, A. Hamann explains how the sharp increase in public deficits during this period crowded out investment and contributed to sluggish growth. He then provides an econometric analysis of the behavior of private savings in Greece, which finds preliminary evidence that policies aimed at increasing government saving are not likely to lead to offsetting movements in private saving.

In addition to crowding out investment and limiting growth, the fiscal profligacy of the 1980s saddled Greece with a huge public debt and rapidly deteriorating debt dynamics, which substantial adjustment efforts in the early 1990s have been insufficient to redress. After a setback in 1993, associated with the electoral cycle, the authorities brought public finances back under control in 1994, and introduced a 1995 budget aimed at a further reduction in the deficit; the debt-to-GDP ratio, however, has continued to rise. Chapter V (prepared by D.G. Demekas) reviews recent developments in public finances and discusses the structure of Greece’s government and government-guaranteed debt. Since in a high-inflation, high-debt country the measured deficit reflects to a large extent the impact of inflation on interest payments, Chapter VI (prepared by A. Hamann) “corrects” the Greek fiscal deficit for this factor and offers some international comparisons of operational (“corrected”) deficits.

Given the size of the imbalance, restoring equilibrium in the Greek public finances is a medium-term undertaking. The authorities’ 1994 revised convergence program (presented in Suppl. 1 to SM/94/151, 7/13/94) aims at a considerable primary adjustment during 1996-98 to reverse the debt dynamics and satisfy the Maastricht convergence criteria in 1998. But the macroeconomic environment has changed since the inception of the program, and the realism of some of its assumptions may be questioned. In Chapter VII, D.G. Demekas presents a medium-term macroeconomic scenario, which suggests that a greater primary effort than that assumed in the convergence program is likely to be needed in order to achieve the program’s overall deficit targets for 1996-98.

For a number of years now, monetary policy in Greece has been gradually shifting from targeting liquidity aggregates to focusing on the exchange rate as a nominal anchor. Recently, the signs of instability in monetary aggregates have multiplied, and the Bank of Greece has placed primary emphasis on the exchange rate (including by officially announcing for the first time in 1995 the targeted rate of crawl vis-à-vis the ECU). This, however, together with the capital account liberalization (completed in 1994), has reduced the room for manoeuver for monetary policy. Chapter VIII (prepared by C. Christofides) discusses the framework and recent conduct of monetary policy in this environment, as well as recent developments in interest rates and the banking system.

In 1994, Greece’s external accounts strengthened considerably: the current account deficit improved to near-balance reflecting strong invisible receipts (mostly tourism), and substantial capital inflows, following the successful handling of the speculative attack against the drachma in May 1994, boosted official reserves to a record US$16.5 billion at end-June 1995. These developments are presented in Chapter IX (prepared by S. Shirai).

The success of the present exchange rate-based disinflation strategy in the coming years depends critically on the maintenance of competitiveness. Chapter X (prepared by S. Shirai) offers an assessment of Greece’s current competitiveness position, looking at a number of indicators, including the exchange rate, export performance, and the overall balance of payments position over the last two decades or so. The risks in such a broad assessment notwithstanding, the Chapter concludes that Greece’s current competitiveness position does not give cause for concern.

Finally, this document includes the usual set of statistical background Tables.

II. Real Sector and Price Developments

1. Aggregate demand, output, and prices

Economic activity recovered somewhat in 1994 (GDP growth is estimated at 1.5 percent) following a mild recession in 1993 relative to the rest of Europe. Moderately stronger growth (of the order of 1.8 percent) is expected for 1995, but it will remain below the European average, (Chart 1) 1/

CHART 1
CHART 1

GREECE Comparisons of Economic Performance

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: WEO; OECD; and Bank of Greece.1/ Average annual percent change in consumer price index.2/ Ratio of gross fixed investment to GDP; data for Greece available only from 1988.

The pick-up in economic activity in 1994 was mainly due to a rise in domestic demand and, to a lesser extent, to a positive contribution of net exports (Table 1 and Chart 2). 2/ Growth in domestic demand stemmed from increases in private consumption (1.5 percent) and investment in machinery and equipment (6.1 percent), with government consumption remaining flat and investment in dwellings contracting (-3.2 percent) for the fourth consecutive year. Both exports and imports of goods and nonfactor services registered positive growth, with the former growing at 7.8 percent, mainly on account of strong tourist activity. Import growth (3.8 percent), on the other hand, was driven by the pick-up in investment in machinery. Overall, net exports’ contribution to GDP growth was 0.3 points of GDP.

Table 1.

Greece: Aggregate Demand

(At constant 1988 prices)

article image
Sources: Ministry of National Economy; and Fund staff calculations.
CHART 2
CHART 2

GREECE Contributions to Growth 1/

(In percent)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: Ministry of National Economy.1/ Total domestic demand is defined as the sum of consumption, fixed investment, and stocks and discrepancies.

The increase in private consumption, largely the result of an increase in real wages (average earnings in the non-agricultural sector grew in real terms for the first time since 1989--see Table 3), was directed mostly to domestic products, thus having a net positive effect on output growth. Government consumption, on the other hand, remained stagnant, reflecting the policy of fiscal retrenchment of recent years: in 1994, government consumption stood 1.6 percentage points of GDP lower than in 1990 (Table 2). The fall in investment in dwellings continued to decelerate in 1994. Negative growth in this category of investment since 1991 seems to be associated with a stock adjustment process triggered by a tightening of tax rules regarding real assets.

Table 2.

Greece: Aggregate Demand

article image
Source: Ministry of National Economy.
Table 3.

Greece: Private Sector Income Account 1/

(In billions of drachmas; at current prices; percentage changes in parentheses)

article image
Source: Ministry of National Economy.

Including public enterprises.

Excluding capitalized interest.

The national saving-investment gap continued to narrow in 1994: foreign saving fell to 0.7 percent of GDP, down from 2.0 percent in 1993 and a peak of 6.5 percent in 1990 (Table 4). About half of the 1.3 percent fall in foreign saving in 1994 can be explained by an increase in national saving (due to a minor improvement in government saving), and the other half by a fall in gross domestic investment. 1/ Private investment grew a little in 1994, after having fallen for three consecutive years, while public investment grew by 2 percent in spite of cutbacks in the public investment budget in that year.

Table 4.

Greece: Savings - Investment Balance

article image
Sources: Ministry of National Economy; and Fund staff calculations.

Different from Table 3 due to inclusion of EU transfers to the public investment budget.

On a national accounts basis; government statistics refer to the general government.

Includes all EU transfers.

As for the sectoral composition of GDP, growth reflects mainly a strong recovery in agriculture (4.5 percent growth) due to favorable weather conditions, and moderate growth in the service sector (1.7 percent) and the manufacturing sector (1.2 percent)--Tables 5 and 6. The manufacturing sector registered positive growth for the first time in 5 years, driven by a growing domestic demand (the sector registered a strong liquidation of stocks in 1994), while high interest rates continued to be a constraining factor (Box 1). Growth in manufacturing was concentrated in the category of non-durable consumption goods, while production of durables remained flat and production of capital goods fell (Table 7). The turnaround in manufacturing performance is also reflected in the industry’s expectations for 1995 (Box 2).

Table 5.

Greece: GDP by Sector of Origin

article image
Source: Ministry of National Economy.
Table 6.

Greece: Agricultural Production

(In thousands of tons)

article image
Sources: Ministry of National Economy; and National Statistical Service of Greece.

Oriental, burley, and virginia varieties.

Table 7.

Greece: Manufacturing Production

(Percentage Changes)

article image
Sources: National Statistical Service of Greece, Monthly Statistical Bulletin; Ministry of National Economy; and IOBE.

Estimate by IOBE.

Factors Constraining Industrial Production in 1994

(percentages of total)

Asked what were the major factors constraining industrial production in 1994, almost one-third of industrial enterprise managers thought that the economic environment imposed no significant constraint on their activity. Of those who responded positively, more than half pointed the finger at high interest rates; only 8 percent of the total thought foreign competition was a problem (Source: ICAP Survey, 1995).

Average inflation was lower in 1994 (10.9 percent) than in 1993 (14.4 percent), but progress in reducing inflation stalled during the year; some deceleration of inflation was nonetheless observed in the first half of 1995 (Table 8 and Chart 3). Year-on-year inflation fluctuated between 10.2 and 11.2 percent in 1994, and reached 10.8 percent at year-end, failing to show a perceptible downward trend. During the first five months of 1995, inflation has decelerated somewhat, but as of June inflation remains only marginally below 10 percent (about three times the EU average). Inflationary expectations remain sticky (Box 3).

Table 8.

Greece: Price Developments

(Average percentage changes over proceding period, except as indicated)

article image
Source: Bank of Greece, Monthly Statistical Bulletin.

Weights are based on 1980 for the wholesale price index, and 1988 for the consumer price index.

This category includes health and personal care and education and recreation, along with other goods and services.

Unit value of imports.

Estimate.

CHART 3
CHART 3

GREECE: Wholesale and Consumer Prices

(Percent changes)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: IMF, International Financial Statistics.1/ Percent change from the same period of the previous year.

Industry Outlook in 1995

(In percent of total responses)

Industrial managers expect a better year in 1995 than in 1994. Two-thirds of them expect sales to increase by at least 10 percent (at or above the expected rate of inflation--see Box 3), and most of the rest expect sales to stay at least unchanged. The outlook for profits is also positive: over 40 percent of managers expect profits to grow in real terms (relative to expected inflation), and another 44 percent expect profits to be at least at their 1994 level (Source: ICAP Survey, 1995).

A breakdown of average inflation by categories (Table 9) reveals a substantial deceleration of inflation in the category “transport and communication”, from 18.8 percent in 1993 to 5.6 percent in 1994. This is partly due to the policy of containing the prices of services provided by public enterprises, following considerable increases in 1993 and earlier. Inflation in the “housing” category also decelerated considerably, from 15.8 percent in 1993 to 10.1 percent in 1994. Inflation in the “food, drink, and tobacco” category, on the other hand, was slightly higher than in 1993, mainly as a result of the increase of taxes on cigarettes.

Table 9.

Greece: Implicit Price Deflators

(Percentage changes)

article image
Source: Ministry of National Economy.

Inflation expectations for 1995

(In percent of total responses)

Enterprise managers are not very optimistic on inflation. The vast majority of them (about 85 percent) expect inflation to be above the official target (8 percent), and almost half expect no deceleration at all relative to 1994, when the inflation rate averaged 10.9 percent. Only a marginal percentage (0.2 percent) expect a drastic decline in the inflation rate, while a substantial percentage are agnostic (Source: ICAP Survey, 1995).

Overall, the effect of cost factors on inflation was mixed in 1994: while prices of imported products increased at a lower pace in 1994 (7.5 percent) than in 1993 (8.6 percent), indirect taxes per unit of output increased at a faster pace and unit labor costs also accelerated somewhat (12.0 percent in 1994 against 11.1 percent in 1993). CPI inflation continued to exceed WPI inflation in 1994, but the gap narrowed temporarily toward the end of 1994, mainly as a result of an acceleration of WPI inflation and the slowdown of inflation in the services sector, which is not captured by the WPI.

2. The labor market

Labor market conditions tightened somewhat in 1994 relative to the previous year. The moderate pick-up in economic activity led to a modest increase in employment--albeit not strong enough to reduce the unemployment rate significantly--and real wages increased for the first time since 1990.

Employment grew by 1.2 percent in 1994, proportional to the expansion of the labor force. As a result, the unemployment rate remained broadly unchanged: 9.6 in 1994 against 9.7 percent in 1993 (Table 10). 1/ Most of the increase in employment was concentrated in urban and semi-urban areas, as employment in rural areas remained stagnant. Relatively strong increases in employment were recorded in the banking sector (7.7 percent) and public enterprises and organizations (4 percent) (Table 11).

Table 10.

Greece: Labor Force, Employment, and Unemployment

(In thousands, unless otherwise noted)

article image
Source: National Statistical Service of Greece.

Period average.

Based on the annual labor force survey by the National Statistical Service of Greece.

By the Labor Force Employment Organization (OAED).

Table 11.

Greece: Employment in Selected Sectors

(In thousands)

article image
Sources: Ministry of National Economy; National Statistical Service of Greece; and Union of Banks.

Seventy thousand persons employed in the repair of vehicles and home appliances have been reclassified into the service sector.

Permanent and temporary employees of the central administration, local authorities, and other budgetary organizations.

Wages in the non-agricultural sector grew in real terms in 1994, following three consecutive years of negative real growth (Table 12). Both the wage bill and average earnings rose faster than inflation, but the level of real wages in 1994 was still slightly below that prevailing in 1988. Wage increases exceeded inflation in manufacturing, retail trade, and the business sector, but fell short of inflation in the civil service.

Table 12.

Greece: Wages and Salaries in the Nonagricultural Sector

(Percentage changes over previous period)

article image
Sources: Bank of Greece; and National Statistical Service of Greece.

National accounts basis.

Differences in rates of change between wage bill and average earnings are due not only to changes in employment, but also to statistical discrepancies.

Gross remuneration (including overtime) in establishments with ten or more employees.

All sectors excluding the civil service, public enterprises, and banking.

Despite rising real wages, unit labor costs (ULC) in the manufacturing sector declined somewhat in real terms in 1994, as productivity per worker increased by 4.1 percent (Table 13). Unlike in previous years, when measured productivity gains stemmed only from reductions in employment, in 1994 average productivity per worker increased also due to higher value added in the sector. While the CPI-based real ULC in 1994 were about 20 percent below their level in 1990, WPI-based ULC adjusted to also reflect increases in employers’ contributions to the social security were slightly above their 1990 level (Chart 4). 2/

Table 13.

Greece: Employment, Productivity, and Unit Labor Costs in Manufacturing

(Annual percentage changes)

article image
Sources: Bank of Greece; and National Statistical Service of Greece.

For wage earners.

Production per manhour.

Estimate.

CHART 4
CHART 4

GREECE Real Wages and Unit Labor Costs in Manufacturing

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: Bank of Greece: and National Statistical Service of Greece.1/ On the basis of CPI inflation.2/ Adjusted for changes in employers’ social security contributions. On the basis of wholesale price inflation.3/ Adjusted for changes in employee social security contributions. On the basis of CPI inflation.

Two-year agreements reached in 1994 provide the basis for contractual wage increases in 1995. In general, nominal wage increases for 1995 will be lower than in 1994: for civil servants, two semi-annual increases of 3 percent each are contemplated, compared with semi-annual increases of 4 percent in 1994; in the private sector, there will be two increases of 4 percent, with some sectors getting an extra percentage point, to be given at an as yet unspecified time, compared with a cumulative increase of 11.8 percent in 1994; in public enterprises, increases of the order of 9-10 percent are expected, as against 14 percent the previous year; and banks will award annual average increases of the order of 10 percent, lower than the 14.3 percent observed in 1994. In addition, in January 1995, civil servants received a 0.8 percent increase resulting from the inflation-correction clause activated when the year-on-year inflation in December 1994 exceeded the threshold of 10 percent stipulated in the agreement. The inflation-correction clause in private sector contracts was not activated, as it was based on a 11 percent threshold, while agreements in public enterprises and banks did not include inflation correction clauses.

The framework for wage negotiations has changed relatively little in recent years, and has proven to be an effective mechanism for achieving a substantial reduction in work hours lost to strikes. In 1994, the number of hours lost to strikes fell to its lowest level since 1980, with sizeable reductions observed in the private sector, as well as in public enterprises and banks (Table 14).

Table 14.

Greece: Collective Labor Agreements, Compulsory Arbitration, and Impact of Labor Disputes

article image
Sources: Bank of Greece; and National Statistical Service of Greece.

Starting in 1992, arbitration decisions are not issued by courts, but by the newly-established (under Law 1876/90) Organization for Mediation and Arbitration.

Various modifications to the existing labor market legislation were introduced in 1994. These include: (i) a law eliminating the provision allowing employers automatically to dismiss workers who participate in illegal strikes, and encouraging dialogue between employers and employees in key public utilities, where a minimum number of workers must continue to work in the event of a strike; (ii) the elimination of various provisions restricting the participation of workers in strikes declared illegal by a court; (iii) a 30 percent increase in unemployment benefits in January 1994, to be followed by an additional 10 percent increase during 1995; 1/ and (iv) increases in contributions to existing funds for training programs, and creation of a new fund to combat unemployment; as a result, non-wage labor costs are expected to increase from their current level of 28.05 percent to 28.50 due to the training fund, and to 28.56 when the unemployment fund becomes operational.

3. Recent developments in privatization and rent controls

After two failed attempts to privatize the National Telecommunications Company (OTE) in 1993 and 1994, the government reduced the share of the company that it intends to put in the market. Currently, the intention is to float about 10 percent of OTE’s shares in the Athens stock exchange in the fall of 1995, perhaps followed by a sale in international markets later (although a decision as to the precise manner in which the stock will be offered in international markets has not been reached). Since preparatory work for the sale of the company has already been made, OTE could be put in the market relatively soon after the final decision is made. The proceeds from the sale of OTE stock are to be distributed between the government (one-third) and the company (two-thirds).

The government has also decided to privatize the oil company (DEP), but has not determined the share of the company to be sold. The actual float is expected to take place by year-end, and the proceeds will go to the holding group. Nevertheless, a debt of Dr 80 billion would have to be paid by the holding group to the central government; of this, Dr 50 billion has already been transferred to the budget in 1994. 1/ No other large public enterprises were considered for privatization.

The state-owned Industrial Reconstruction Organization (IRO) privatized seven commercial or industrial firms in its portfolio in 1994, and four in the first four months of 1995. 2/ The last seven firms currently in operation (employing some 4,000 workers) are to be privatized, and 52 firms operating only with skeleton personnel are in the process of liquidation. In addition, the Agricultural Bank and the National Bank sold six firms in 1994. Since last year, no ailing firms have been added to IRO’s portfolio, reflecting the government’s policy of allowing these firms to either file for bankruptcy or be liquidated.

Upon expiration of the regime of rent controls for residential property in September 1994, the government introduced a set of measures amounting to a gradual liberalization of rents:

  • As of September 1994, all rents of units occupied for more than 12 years, and units occupied for at least 9 years and larger than 90 sq. m. were liberalized.

  • As of July 1995, all rents of units occupied for at least nine years irrespective of size, as well as units occupied for at least five years and larger than 90 sq. m. are to be liberalized.

  • As of July 1996, all rents of units leased for at least four years are to be liberalized.

The dates refer to the time when expiring leases will have to be renegotiated. However, eviction procedures cannot start until six months later. All new rental agreements would have to be signed for periods of at least three years; tenants could, however, break the lease.

As for commercial property, the government enacted as law the three-year rent decontrol program described in last year’s Recent Economic Developments document (SM/94/173). Accordingly, rents in properties that have been rented for more than 30 years are to liberalized in September 1995; units rented for more than 20 years, in September 1996; and units rented for more than 10 years, in September 1997.

III. An Index of Coincident Economic Indicators for Greece

Economic policy making in Greece is hampered by the absence of an up-to-date, high-frequency indicator of overall economic conditions: quarterly estimates of gross domestic product (GDP) are produced with a considerable lag and are of doubtful quality; annual GDP estimates are also produced with a considerable lag. In practice, of course, policy makers have at their disposal a number of different high-frequency indicators, but there is no information available regarding the ability of such indicators to track overall economic conditions.

This Chapter reviews the existing methodologies for constructing indices of coincident economic indicators; examines the suitability of several existing high-frequency (monthly) indicators for constructing such indices; and constructs a number of alternative coincident indices for Greece, and evaluates their performance. On this basis, this Chapter provides estimates of the current economic activity in Greece based on the latest available monthly indicators. These suggest that output growth by end-1994 was about 2 percent, higher than the annual GDP growth rate of 1.5 percent estimated by the authorities.

1. Methodologies for constructing a coincident index of economic activity

There are two broad approaches for constructing coincident indices: an older methodology, which is still used extensively by government agencies and other institutions, and a newer set of methods proposed by academic researchers, which utilize new time series techniques.

The older methods go back to the contributions of Mitchell and Burns (1938), and have changed relatively little since then. 1/ The basic idea is that business cycles are reflected in a number of different monthly indicators, which fall into three categories: leading, coincident, and lagging. “Reference dates” are determined by looking at turning points for many different series, which are then classified accordingly. The series in each category are then weighted according to a number of criteria 2/ and, after some further manipulations, such as de-trending, smoothing, and scaling, the weighted series are aggregated to produce a single monthly index. Variants of this system of producing indices are used by the Bureau of Economic Analysis (BEA) of the U.S. Commerce Department, 3/ the National Bureau of Economic Research, the OECD, the U.K. Central Statistical Office, 1/ and in many other countries. 2/

The popularity of the older methods attests to their usefulness and practicability. These methods have been criticized, however, from the point of view of modern time series econometric methodology. Some criticisms center on the arbitrary way of calculating weights for the component series and the related issue of which series to include. 3/ Another problem has been that it is not obvious how to interpret swings in the index: is a one month change enough to signal a turning point? or is a longer time horizon required? A more fundamental criticism has been that the procedure for constructing the indices does not allow their statistical properties to be determined.

This last criticism is the starting point for the newer methods proposed by Stock and Watson (1988, 1993), henceforth referred to as SW. SW begin by assuming that there exists a single unobservable factor, the “state of the economy”, which affects other observable series. By assuming an explicit time series representation of the unobservable variable in terms of observable series (technically, a dynamic factor analysis or single index model), and a dynamic representation for the unobservable variable, SW are able to estimate the “state of the economy” for the United States (the coincident index) using the Kalman filter. This technique allows for a formal derivation of the statistical properties of the index.

An important problem of both the older and the newer methods of constructing coincident indicators is the difficulty of identifying a benchmark for validating the results, since both methods estimate an “unobservable”. SW provide statistical tests of some of their assumptions, such as whiteness tests on the residuals of the observable series and integration/cointegration tests on the observable series. These tests provide support for the internal coherence of the model that they propose; however, they cannot by themselves determine the accuracy of the resulting index. For this, SW compare their index to the BEA index: they find that the two indices, despite being computed with different methods, are very similar. But if the older indices can be used as a benchmark for the newer methods, what can be used as a benchmark for the older indices? One “natural” candidate would be a GDP series, but it is not available on a monthly basis. 4/ Thus, in practice, industrial production is often used on the grounds that it captures well the turning points in business cycles. But this argument has problems: first, the coincident indices usually include averages which incorporate industrial production, so it is not surprising that they tend to track industrial production well; and, second, if industrial production is a good indicator of turning points in activity, why not simply use it as a coincident indicator?

The approach followed here for constructing a coincident index for Greece is to use GDP as a benchmark. Even if, as SW assert, the concept of business cycle is broader than GDP, estimating a coincident index that tracks GDP would be very useful for policy. Moreover, this approach has two advantages: (i) GDP is partially observable, and this provides a method of validating the results; and (ii) a coincident index that directly tracks GDP is superior to the usual coincident indicators which merely track turning points without indicating the strength of the upswings/downswings. 1/ The proposed methodology is described below.

Consider a business cycle variable yt, which in what follows will be taken to refer to GDP, that is observable at some relatively low frequency, such as quarterly or annually. This variable is decomposed into higher frequency variables, to be denoted by xt, which are unobservable:

y t = x t + x t − 1 + x t − 2 + … + x t − ( k − 1 ) ( 1 )

When y is observed annually and x is quarterly, k=4; for x monthly, k=12. Using the lag operator L, where for any time series zt, Lzt = zt-1:

m k ( L ) = 1 + L 1 + L 2 + … + L k − 1 ( 2 )

With this notation, equation (1) can be written compactly:

y t = m k ( L ) x t ( 3 )

Although observations of xt are not available, xt is statistically related to a potentially large set of observable series, collected in vector zt. Thus,

A ( L ) x t = B 0 ( L ) z t + C 0 ( L ) u t ( 4 )

where A(L), B0(L), and C0(L) denote appropriately-dimensioned polynomials in the lag operator and ut denotes a white-noise error term. Albeit linear, the time-series representation posited here is quite general in that it allows all possible lagged effects. However, for zt to be used as a coincident (rather than leading, or lagging) indicator, most of the explanatory power should be found either contemporaneously or with relatively short lags.

If xt is stable and does not have a unit root, it is possible to derive a reduced form of equation (4):

x t = B ( L ) z t + C ( L ) u t ( 5 )

where B(L) = A-1(L)B0(L) and C(L) = A-1(L)C0(L), or by substituting xt from equation (3):

y t = B ( L ) [ m k ( L ) z t ] + C ( L ) [ m k ( L ) u t ] ( 6 )

If, on the other hand, xt has a unit root, a representation akin to equation (5) exists for xt, which is differenced to remove the unit root:

Δ x t = B ( L ) z t + C ( L ) u t ( 7 )

where Δ = (1-L). In terms of the year-on-year change 1/ of yt (or, when yt is measured in logarithms, the year-on-year growth rate):

Δ k y t = B ( L ) [ m k 2 ( L ) z t ] + C ( L ) [ m k 2 ( L ) u t ] ( 8 )

where ΔK = (1-LK) and mk2(L) is the square of the moving sum operator. This can be seen by noting that ΔKyt = mk(L)Δyt and Δyt = mk(L)Δxt, substituting for Δxt from equation (6) and rearranging.

The important characteristic of both (6) and (8) is that they link observables: even though yt and Δkyt are not observed at all times, they are partially observed. Equations (6) and (8) can therefore be estimated using one of the conventional methods that deal with missing observations. Diagnostic statistics (such as R-square, t-statistics, and the Durbin-Watson statistic) can then provide a direct test of overall fit and statistical significance. This is the sense in which the partially observed nature of GDP can serve to validate the model. The same diagnostic statistics can indicate whether equation (6) or equation (8) is superior. 2/ Once the coefficients of the lag polynomials B(L) and C(L) are estimated, the coincident indicators zt, which are observed at all times, can be used to provide estimates of the moving sum of GDP even when GDP is not observed.

If desired, the estimated coefficients in B(L) and C(L) can be used to generate estimates of xt--i.e., monthly estimates of GDP that are as current as the coincident indicators in zt. Of course, such estimates will in general be move volatile and subject to wider margins of error.

2. Available Greek data for the construction of a coincident index

Greece produces a number of monthly series similar to the set used by the BEA that can serve as coincident indicators. The most relevant are:

article image

A first inspection of the series reveals that a good number appear to track well the path of GDP (Chart 5). The industrial production series seems to track GDP particularly well, although its path seems to have diverged somewhat from that of GDP after 1988. The retail sales series behaves in a similar way, except that it is somewhat more volatile. There is also some indication that a portion of retail sales might have recently began to be distributed through channels that are not captured by the retail sales index. The production of cement series seems to track GDP very well, especially after 1979. Finally, the narrow real money supply series tracks GDP well only after the mid-1980s. The other four series do seem to have useful information, but each suffers from some deficiency (Chart 6). New passenger car registrations are very volatile and subject to long swings. Employment in manufacturing has diverged very significantly from GDP after 1987 and seems influenced by long-term, secular productivity increases or shifts in the share of manufacturing in the economy. And imports/exports reflect the increasing openness of the Greek economy. To summarize, the first four series seem the most promising candidates for inclusion in the coincident index, although the remaining four series should also be tested for information content.

CHART 5
CHART 5

GREECE Coincident Indicators I

(Logarithmic index, end-1985 equals zero)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Greek authorities and Fund staff estimates.
CHART 6
CHART 6

GREECE Coincident Indicators II

(Logarithmic index, end-1985 equals zero)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Greek authorities and Fund staff estimates.

With these series, a number of different coincident indices can be constructed. Their timeliness, however, depends on the timeliness of the series used for their construction, as indicated below.

Given the timeliness of the available series as of May 1995, the coincident indices listed above can provide an indication of the state of the Greek economy for up to end-1994.

Alternative Coincident Indices

article image

Good indicators have a lag of 4-5 months. Mediocre indicators are three months behind good indicators. Bad indicators lag mediocre indicators by a further four months.

3. Constructing a coincident index for Greece

Utilizing the methods and data previously described, a number of coincident indices were constructed for Greece. The table below provides preliminary estimates and diagnostic tests.

The results suggest that in general, indices with broader coverage are better, since they tend to have lower standard errors. However, this gain comes at a steep cost, since the broader-based indices are less timely. Examining the levels-based coincident indices, it is found that the industrial production index tends to have the greatest weight, followed by retail sales. Production of cement is a useful predictor, but only in the broader-based indices. Conversely, money tends to be useful for capturing information for variables that are missing from the narrower index. Employment is an important variable, although it has a negative sign. This result perhaps means that employment is capturing productivity effects in the short run: lower employment is related to a higher productivity, which is positively related to output. Finally, both car registrations and exports seem to have predictive power, while imports do not: perhaps the informational content of imports is already captured by other series, notably retail sales.

Regression Results for Alternative Coincident Indices 1/

article image

Explanatory variables measured as 12-month moving sums for coincident indicators based on level models and as 12-month moving sums squared for difference models.

The unit root-based coincident indices seem to perform less well, judged by the larger standard errors of the estimates. The coefficient for industrial production is counterintuitive, although retail sales and car registrations retain some predictive power. In effect, the assumption of a unit root imposes a lot of structure on the model: this is helpful if the assumption is correct, but damaging if the assumption is incorrect. If the Dickey-Fuller statistic reported earlier is to be believed, the unit root assumption is incorrect.

Comparing the performance of the two narrower indicators (XL1 in Chart 7 and XD1 in Chart 8), chosen on the basis of their superior timeliness, shows that the levels-based indicator has a lower standard error (a measure of a one-step ahead forecast error), and seems better to track actual GDP. The unit root method appears to impose too much smoothness on the coincident index, which would be what one would expect if the unobserved index did not exhibit the same degree of persistence implied by a unit root process. This comparison seems to support the levels-based indicator XL1.

CHART 7
CHART 7

GREECE XL1: Actual vs Estimated Output

(Year-on-year logarithmic change)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Greek authorities and Fund staff estimates.
CHART 8
CHART 8

GREECE XD1: Actual vs Estimated Output

(Year-on-year logarithmic change)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Greek authorities and Fund staff estimates.

Based on XL1 estimates, the business cycle in Greece appears to have turned around at end-1993. The year-on-year rate of growth reached around 1.5 percent by mid-1994, higher than the 1 percent annual rate of increase predicted by the authorities at about that time. By year-end, the rate of increase exceeded 2.0 percent (Chart 9), higher than the 1.5 percent currently predicted by the authorities (although the 1.5 percent annual rate is well within one standard deviation of the XL1-estimated rate of growth).

CHART 9
CHART 9

GREECE XL1: Actual vs Estimated Output (level)

(In real 1985 drachma billion)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Greek authorities and Fund staff estimates.

IV. Private Saving Behavior in Greece

1. Introduction

This chapter analyzes the behavior of private saving in Greece, and in particular the role of fiscal variables in the determination of private saving. The importance of studying saving behavior in Greece is highlighted by the high correlation found between domestic investment and national saving over the last 30 years. For small economies such as Greece, this result has been shown to reflect either low international mobility of long-term capital or the impact of policies aimed at preventing the current account of the balance of payments from departing substantially from balance. 1/ In either case, the sustained increases in investment needed to enhance the economy’s growth potential (such as that envisaged in Greece’s medium-term convergence plan) cannot materialize without commensurate increases in national saving.

The chapter provides preliminary evidence that in Greece fiscal policy can play a crucial role in achieving a sustained increase in national saving and, in particular, that policies aimed at increasing government saving are not likely to lead to offsetting movements in private saving. On the one hand, private saving is shown not to be affected by changes in current taxes, probably due to widespread liquidity constraints--suggesting that an increase in tax revenues would raise government and national saving by the amount of the additional tax revenue. On the other hand, government consumption is shown to be negatively correlated with private saving, reflecting complementarity between private and government consumption, and implying that increases in government saving achieved through reductions in government consumption “crowd-in” private saving and lead to even higher increases in national saving.

These results for Greece differ from various empirical findings of some degree of substitutability between private and government saving, but are consistent with previous work on the effects of fiscal policy on private saving in OECD countries, and with studies on the relationship between private and government consumption both in Greece and in other countries.

The remainder of this Chapter is organized as follows. Section 2 compares the main trends in investment and saving in Greece with those observed in other industrialized countries and discusses the importance of national saving; section 3 describes the behavior of inflation-adjusted private and government saving rates; section 4 discusses the results of the estimation of an econometric equation for private saving in Greece; and section 5 draws the main conclusions.

2. Long-term trends in saving and investment

Several studies focusing on trends of investment and saving in industrial countries have consistently found the same stylized facts: (i) a relatively high correlation between domestic investment and national saving; and (ii) a marked decline in national saving rates during the 1970s and 1980s, related to growing fiscal imbalances. These findings suggest that fiscal policy can have an important effect on investment through its impact on national saving.

a. The correlation between saving and investment

Several explanations have been proposed for the high correlation found between domestic investment and saving. In a popular study, Feldstein & Horioka (1980) (F&H), after controlling for various factors that could render both saving and investment endogenous, suggest that the high correlation between them reflected low international mobility of long-term capital. Other researchers have questioned this conclusion: Fieleke (1982) and Bayoumi (1990), among others, argue that the correlation arises as a result of policies aimed at keeping the current account of the balance of payments near balance. Murphy (1984), on the other hand, provides an additional explanation for the high correlation between domestic saving and investment, which applies only to large countries: even under high capital mobility, the ability of large countries to affect the level of world interest rates would result in domestic saving and investment to move in tandem over time. For example, a surge in saving in the United States would result in lower world interest rates, which would thus stimulate investment world-wide, as well as in the United States. 1/

Does the stylized fact of a high correlation between domestic saving and investment also appear in Greece? The Table below shows the following results for EU countries: 2/ (i) simple correlation coefficients between national saving and investment rates; (ii) OLS estimates of regressions of the investment rate on savings; and (iii) information on the relative size of the countries in the sample.

Correlation Between Saving and Investment Rates in the EU

article image
Sources: Lemmen and Eijffinger (1993), and IMF WED data base. An asterisk (*) indicates significance of the slope parameter at the 5 percent level.

Average GDP of country i divided by the average GDP of the OECD. All figures were converted to U.S. dollars. Averages were computed over the period 1970-89, the same period used for the estimations.

Two groups can be distinguished: a group of eight countries with high correlations between saving and investment and relatively high and statistically significant regression coefficients, and a group of four countries with low correlation and statistically insignificant coefficients. In the first group only Germany, France and Italy could be considered large enough to suspect that the correlations reflect mainly country size considerations, whereas for Greece, Denmark and Belgium, all small countries, the high correlations are likely to reflect either low international mobility of long-term capital and/or policy effects.

b. Rising fiscal imbalances and declining saving and investment rates

Saving and investment rates in most industrial countries have declined from the high levels observed in the early 1970s, and in most cases the decline has roughly coincided with increasing deficits in their government current account balances. This phenomenon has been pointed out in many studies--see, for example, Aghevli et al. (1990) and more recently IMF (1995)--and is illustrated for the EU in Chart 10.

CHART 10
CHART 10

GREECE European Union: Investment and Saving

(In percent of GDP)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: IFS.

The top panel of Chart 10 shows EU-wide investment and national saving rates since 1970. Both rates declined substantially during the 1970s and early 1980s and, after recovering somewhat at the end of the 1980s, both fell again in the 1990s. In 1993, national saving and investment rates stood at their lowest level in the period. The bottom panel reveals a close relationship between national and government saving rates. Government saving also deteriorated substantially during the 1970s, turning negative in 1981, recovering somewhat in the late 1980s, and falling again in the 1990s. It is noteworthy that throughout the period, the private saving rate remained relatively stable hovering at about 20 percent of GDP.

The comparison of average levels of saving and investment rates during the 1970s, 1980s, and early 1990s illustrates the close link between them. The EU-wide national saving rate averaged 24.4 percent in the 1970s, 20.8 percent in the 1980s, and 19.9 percent in the period 1990-93. The average for the government saving ratio was 2.7 percent in the 1970s, zero in the 1980s, and -1.4 percent in the period 1990-93. Thus, three quarters of the decline in the average national saving rate between the 1970s and the 1980s and almost all of the decline between the 1970s and the early 1990s can be attributed to the behavior of government saving.

Chart 11 (top panel) shows similar patterns for national saving and investment rates in Greece. 1/ Both rates began falling in the 1970s and, while investment recovered somewhat in the late 1980s but declined again in the 1990s, saving registered a modest recovery in the 1990s. Movements in national saving also seem to be closely related to movements in government saving in Greece, but a detailed description of the breakdown of national saving into private and government is left to the next section, where the national account figures are adjusted for the effects of inflation.

CHART 11
CHART 11

GREECE Investment, Private Saving and Government Saving

(In percent of GDP)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Data provided by the authorities; and staff calculations

3. Inflation-adjusted saving rates in Greece

National accounts in countries with high public debt ratios and relatively high inflation rates, such as Greece, are generally subject to large distortions introduced by inflation. The bias is due mostly to the classification of the entirety of interest payments on domestic debt as disposable income of the private sector and as current expenditures of the government. The bias can not only amount to several points of GDP if both inflation and the domestic currency-denominated debt are high, but it can also distort national accounts to a varying extent if inflation itself fluctuates considerably, as has been the case in Greece over the last 30 years. Thus, it is essential to correct the Greek figures of private and government saving for the presence of inflation in order to obtain measures of saving flows that better resemble their theoretical counterparts.

A detailed explanation of the inflation adjustment can be found in the annex to this paper. In general, the adjustment consists of a computation of the erosion by inflation of the stock of government bonds and base money held by the nongovernment sector. The resulting term is subtracted from private saving and added to government saving. As a result, inflation adjustment changes the composition of national saving between government and private but not its size.

The middle and bottom panels of Chart 11 present the breakdown of national saving into private and government saving using conventional and inflation-adjusted figures, respectively. Clearly, inflation adjustment reduces the differences between private and government saving rates, and eliminates a slight upward trend in the private saving rate, but it does not alter the main conclusion: that the decline in national saving rates was, to a great extent, the result of the deterioration in the fiscal position, especially since the 1980s.

Examination of the middle panel of Chart 11 reveals a high correlation between the national and government saving rates, but also a slight upward trend in private saving. The national saving rate averaged 26.1 percent in the 1970s, 19.2 in the 1980s, and 18.3 in the period 1980-1993. According to the conventional figures, the reduction of 6.9 percentage points in the national saving rate between the 1970s and the 1980s can be broken down into a fall of 9.3 points in government saving and an increase of 2.4 points in private saving. Between the 1970s and the 1980-93 period, the fall in the national saving ratio was 7.8 percentage points, resulting from a fall of 10.3 points in the government saving rate and an increase of 2.5 points in the private saving rate. The conclusion is that a sizable fall in government saving drove down the national saving rate, although it was partly offset by a modest increase in private saving.

When inflation-corrected figures are used, variations in the national saving rate can be almost fully ascribed to movements in the government saving rate, with no offsetting movements in private saving. The 6.9 percentage point fall in the national saving rate between the 1970s and 1980s can be broken down into a fall in government saving of 6.1 points and a fall of 0.8 point in the private saving ratio. If the comparison is made against the 1980-1993 period, the fall in the government saving ratio is 6.7 points, out of a total fall of 7.8 points. These results show that the behavior of the national saving rate can be explained almost exclusively by the evolution of inflation-adjusted government saving. The next section provides a more detailed econometric analysis of this result.

4. A private saving equation for Greece

a. The determinants of private saving

The role of a set of variables in the determination of private saving in Greece is examined beginning with the following simple equation for the private saving rate:

S P t = β 0 + β 1 T t + β 2 G t + β 3 E t + β 4 R t + β 5 Y t + μ t ( 1 )

where SP, T, and G represent private saving, taxes, and (general) government consumption as a percentage of GDP; E is the ratio of individuals of ages 65 and over in the total population; R is the real interest rate defined as the difference between the nominal interest rate on 3-12 month deposits at commercial banks and the actual inflation rate; and Y is the log of real GDP, included in order to account for the possibility of a non-unit income elasticity of real private saving.

The effects of fiscal policy on private saving are assessed on the basis of two variables: taxes and government consumption. The effect of changes in taxes for a given level of government consumption is captured by β1. If Ricardian equivalence held, increase in taxes without a corresponding change in government consumption would not have any effect on private consumption, and would thus reduce private saving only to the extent that it reduced disposable income; consequently, β1 would be -1. On the other hand, if liquidity constraints or other factors making consumption highly sensitive to disposable income prevailed, changes in taxes would have little or no effect on private saving, as disposable income and private consumption would tend to move together; in such a case, β1 would be close to zero.

The effect of changes in government consumption are a priori more difficult to interpret. In a Barro-type framework, a permanent increase in government consumption with no current changes in taxes would lead to a 100 percent crowding out of private consumption, an equivalent increase in private saving, and, thus β2 would be 1. But the outcome could be less than-full crowding out under different assumptions--e.g., if the increase in government consumption was temporary. The same result could obtain in models with finite lifetimes and no altruism, or with imperfect capital markets. In these cases, β2 would be positive but less than 1.

In addition to uncertainty about whether changes in government consumption are temporary or permanent, there are other reasons for ambiguity in the magnitude, and even the sign of β2. Government consumption and private saving could be directly linked to each other if the marginal utility of private consumption is affected by government consumption. In the traditional scenario discussed by Bailey (1971), private and government consumption are substitutes and, therefore, an increase in government consumption would lower the marginal utility of private consumption and lead to a reduction in it and, thus, to an increase in private saving. In this case, β2 would also be positive, but its value would depend on the degree of substitutability between private and government consumption.

However, the substitutability hypothesis has found little support empirically. Using data for the United States, Aschauer (1985) found a low degree of substitutability between private and government consumption, while Karras (1994), using a sample of 30 developed and developing countries, found that private and government consumption are complementary in most cases. Under the complementarity hypothesis, increases in government consumption would increase the marginal utility of private consumption and lead to higher consumption and lower private saving, implying a negative sign for β2.

The issue of substitutability or complementarity of private and government consumption is further complicated by the fact that different categories of government consumption may affect the marginal utility of private consumption in different ways--Cerea (1982). A priori, government spending on “public goods” such as defense, law and order, and public administration, tends to raise the marginal utility of consumption and, thus acts as complementary to private consumption. On the other hand, “merit goods”, such as publicly-provided health and education, can be a direct substitute of private consumption.

The effects of changes in the share of the elderly in the total population is captured through β3. The ratio is a purely demographic variable, not an exact measure of the dependency ratio, and is intended to capture the effect that an increase in the average age of the population has on the private saving rate. The life-cycle theory of saving would predict a negative effect, since the elderly tend to dissave, consuming out of the assets accumulated during their working years.

Conceptually, the effect of the real interest rate on saving is ambiguous: the positive substitution effect of an increase in the real interest rate may be offset by the negative effect that a higher real interest rate could have on the present value of the future stream of income. Empirically, the results have indeed been mixed, and in many instances the real interest rate has not been found to be a significant regressor in saving equations. 1/

Before turning to the estimation of the saving equation, it should be pointed out that a popular way of assessing the effects of fiscal policy on private saving uses the government deficit (or government saving) as an explicit regressor, and associates Ricardian equivalence with the existence of a coefficient of -1 for the deficit variable. Equation (1) discussed above can be re-written in order to allow for the government balance to appear as a regressor: 2/

S P t = β 0 + β 1 ( T t − G t ) + ( β 1 + β 2 ) G t + β 3 E t + β 4 R t + β 5 Y t + μ t ( 2 )

In this formulation, the coefficient of the surplus variable β1 is the same as the coefficient of the tax variable in equation (1), and measures only the effects of changes in taxes (changes in the deficit not related to changes in government consumption). Thus, it can be used to measure the degree of offsetting (Ricardian equivalence). The coefficient of government consumption is now (β1+β2) in terms of the parameters of equation (1), and measures the effect on private saving of tax-financed increases in government spending (increases in government consumption leaving the deficit unchanged). Equations (1) and (2) are identical, and the rearranging of terms implicit in equation (2) does not add any information. The results presented below are discussed in terms of equation (1).

b. Estimation

Three versions of the private saving equation discussed above were estimated: (i) the static model represented by equation (1); (ii) an unrestricted dynamic model including one lag of each of the variables in (1); and (iii) a restricted dynamic model, in which non-significant terms in the unrestricted dynamic model were dropped. The results are shown in the table below. In addition, in order to ensure that the possible nonstationarity of at least some of the variables involved has not led to spurious correlations, equation (1) was re-estimated using Johansen’s cointegration technique.

All estimated parameters in the static model were found to be statistically significant, and the non-fiscal variables had the expected signs. Their estimated values were somewhat surprising, however, especially in the cases of the elderly ratio and real income. An increase of one point in the elderly ratio was found to lead to a reduction of almost 6 percentage points of GDP in the private saving ratio, a highly implausible result. An increase of 1 percent in real GDP is estimated to raise the private saving ratio by about 0.37 percentage points of GDP, suggesting a high income elasticity for private saving. The real interest rate, on the other hand, was found to have a positive but relatively weak effect on private saving: an increase of one percentage point in the real interest rate raises the saving ratio by about 0.13 points of GDP, about one half the magnitude estimated for a group of 21 industrial countries in IMF (1995).

Private Saving Equations 1/

article image

All equations estimated using OLS. An asterisk denotes significance at the 5 percent level.

Except for the long-run coefficients, for which a Wald test of the null hypothesis that the sum of the short-run coefficients equals zero is reported.

The estimated parameters for the fiscal variables are also somewhat surprising: the coefficient of the tax variable is -1, indicating perfect Ricardian equivalence. Most studies on private saving find only partial offsetting (about one-half in IMF 1995), and a higher offsetting coefficient in a country in which liquidity constraints are likely to be substantial is quite improbable. The sign of government consumption is negative suggesting complementarity between private and government consumption.

The long-run parameters obtained from the two dynamic formulations are quite different, suggesting the possibility that the estimates obtained from the static model were subject to misspecification bias. In the unrestricted dynamic model, the long-run impact of an increase in the elderly ratio is reduced by about one-half, but becomes only marginally significant (the significance of the t test is 15 percent). The effect of changes in real income is reduced by about one-third and remains significant. The effect of the interest rate, on the other hand, appears stronger. The long-run coefficient of government consumption is close to -1 indicating almost perfect complementarity between private and government consumption, while the short- and long-run coefficients of the tax variable become statistically insignificant, ruling out any degree of Ricardian equivalence.

The restricted dynamic model yields a stronger negative effect of government consumption on private saving (although the coefficient is not statistically different from -1), a higher real interest rate semi-elasticity (comparable to the one estimated by IMF 1995), and a lower long-run effect of changes in real income. However, the long-run effect of changes in the elderly ratio becomes statistically insignificant.

In order to deal with the possibility of spurious correlation due to the non-stationarity of at least some of the variables used in the estimations, the long-run parameters of the private saving equation were re-estimated using Johansen’s cointegration method; the results are shown below. The top panel of the table shows the results of Johansen’s test for the number of cointegrating vectors: both the maximum eigenvalue and the trace tests are indicative of the existence of 2 cointegrating vectors. The signs and magnitudes of the parameters estimated in the first cointegrating vector (shown in the bottom part of the table) suggest that it can be interpreted as a long-run private saving equation. They are quite similar to the long-run parameters obtained from the dynamic equations. Moreover, statistical tests of the significance of the parameters confirm the result obtained with the dynamic models: (i) that taxes do not play a significant role in the determination of the private saving ratio; (ii) that government consumption is (strongly) negatively correlated with private saving; (iii) that the elderly ratio is significant only at the 15 percent level; and (iv) that the real interest rate and real income variables are statistically significant and have the expected signs.

Private Saving Equation: Cointegration Anaylsis 1/

article image

One asterisk denotes significance at the 5 percent level. Two asterisks denote significance at the 1 percent level.

First cointegrating vector.

Zero restriction test on the coefficient of the respective variable on both eigenvectors (test performed with rank = 2).

Significant at the 15 percent level.

In sum, the results discussed above can be interpreted as follows. First, taxes do not seem to play a role in the determination of private saving, perhaps reflecting a high degree of dependence of current consumption on disposable income. This implies that increases in government saving achieved through higher taxes will not lead to any offsetting behavior by the private sector, and will thus lead to a one-to-one increase in national saving. Second, increases in government consumption have a substantial negative impact on private saving, indicating a high degree of complementarity between private and government consumption and implying that the government has a powerful tool to affect overall national saving, because a reduction in government consumption would not only increase government saving but would also lead to higher private saving.

The results obtained for the effects of fiscal policy on private saving, while in contrast with some empirical findings of a degree of substitutability between private and government saving, are nonetheless consistent with findings of some studies for industrial countries, and for Greece in particular. Using a sample of OECD countries over the period 1972-1987, McCallum (1992) found a strong negative correlation between government spending and private saving. Tridimas (1992) examined the relationship between different categories of government spending and private consumption in Greece over the period 1958-1987 and found strong evidence of complementarity between spending on “public goods” and private consumption, but not a significant relation between spending on “merit goods” (health and education) and private consumption. With consumption on public goods about 80 percent of government consumption in Greece, this result is consistent with the negative relationship between government consumption and private saving reported above.

5. Conclusions

The behavior of saving, both private and government, and investment in Greece over the last 30 years seems to indicate that, due to low international mobility of long-term capital, or government policies aimed at preventing sustained disequilibria in the current account of the balance of payments, national saving is an important determinant of domestic investment. The finding that fiscal policy can have strong effects on national saving means that the government can effectively promote investment and potential growth in Greece. Although policies aimed at raising government saving through tax increases lead to comparable increases in national saving due to the absence of an offsetting effect in private saving, reductions in government consumption are shown to be a more powerful tool to raise national saving.

The results of the estimations carried out here should be taken carefully, because they are subject to the usual measurement and interpretation problems, especially with respect to the effect of the fiscal variables (see Seater 1993). Additional work is needed in order to explore in more detail the link between private and government consumption, and to test other specifications of the saving equation. Overall, however, the results seem to be quite robust and to provide strong support to the conclusions drawn above.

Finally, it should be pointed out that recently enacted financial liberalization measures are not likely to lead to a higher degree of long-term capital mobility in the near future and, thus, do not weaken the conclusion that higher national saving is needed in order to sustain an increase in investment rates over the medium term. Moreover, the case for increased national saving is strengthened by the government’s objective of keeping the current account near balance.

ANNEX: Inflation Correction of Saving Data in Greece

This annex describes the methodology used to adjust the Greek national account figures of private and government saving for inflation. The need for this correction arises from the fact that the erosion of claims against the government held by the private sector are not properly accounted for in the national accounts. On the one hand, part of the government outlays commonly registered as interest payments on domestic debt actually represent amortization on that debt, which biases government saving downward. On the other hand, erosion by inflation of the monetary base, which simultaneously reduces the real value of a government liability and a private asset, has no “above the line” counterpart in the national accounts. As a result of both phenomena inflation leads to overestimation of private disposable income and private saving, and an underestimation of government saving. 1/

Further refinements of the data in order to properly account for other factors unaccounted for by the national accounts aggregates such as saving through purchases of durable goods, capital gains and losses due to exchange rate changes, among others would certainly improve the quality of the data, but was judged unfeasible given the limitations of the available information.

In order to correct the distortions created by inflation, an “inflation correction term” (ICT) is computed and then simultaneously added to government saving and subtracted from private saving:

SP ( adj ) = SP − ICT SP ( adj ) = SP + ICT A 1

The adjustment term is the sum of two components: the net inflation tax on base money (ITMO) and the gross inflation tax on other government liabilities (ITG):

ICT = ITMO + ITG ( A2 )

The first term was estimated as the inflation erosion of the monetary base (MO) net of interest payments on bank reserves (R) and private deposits at the Bank of Greece (DP), at the rates iR and iD respectively.

ITMO = π MO − [ I R R + i D DP ] ( A 3 )

The second term was computed as the product of the inflation rate and the sum of T-bills held by the public (TBP), T-bills held by banks (TBB), and other bonds and government securities held by banks (GBB): 1/

ITG = π [ TBB + TBP + GBB ] ( A4 )

The following definitions were used to compute the inflation adjustment terms using Greek data: (i) the annual inflation rate was computed as the percentage increase in the average CPI; (ii) iR was assumed to be zero throughout the period 1964-1993; and (iii) iD was proxied by the average annual rate on deposits (3 to 12 months) at commercial banks. In all cases, beginning-of-period stocks of base money or other financial assets were used in the computations.

V. Public Finances and Debt

1. Overview of developments during the early 1990s

After a continuous deterioration of the public finances during the 1980s, an effort was made to restore equilibrium in the early 1990s. During 1990-92, a cumulative primary adjustment of some 9 percentage points of GDP took place, turning the central government primary balance from a deficit of 7½ percent of GDP in 1989 to a surplus of 1.4 percent in 1992, and reducing the overall deficit from 14.3 percent in 1989 to just under 10 percent in 1992 (Tables 15 and 16). This adjustment, however, was not sufficient to prevent further increases in the stock of debt: at end-1992, central government debt reached about 94 percent of GDP (including military debt).

Table 15.

Greece: Summary of Public Finances

(In billions of drachmas: accrual basis)

article image
Sources: Ministry of National Economy; and Bank of Greece.

Does not include Dr. 462 billion (2 percent of GDP) of interest payments due on December 31, 1994 (but payable on January 2, 1995), which are, however, included in the provisional outcome for 1994.

Special Account for Guarantees of Agricultural Products.

Bank of Greece data.

Staff estimates; not comparable to official data on a national accounts basis.

Table 16.

Greece: Summary of Public Finances

(In percent of GDP: accrual basis)

article image
Sources: Ministry of National Economy; and Bank of Greece.

Does not include Dr. 462 billion (2 percent of GDP) of interest payments due on December 31, 1994 (but payable on January 2, 1995), which are, however, included in the provisional outcome for 1994.

Special Account for Guarantees of Agricultural Products.

Bank of Greece data.

Staff estimates; not comparable to official data on a national accounts basis.

The increases in the debt were driven by the budgetary effects of financial liberalization, which started in the late 1980s. These resulted in: (i) increased transfers and assumption of nonperforming assets of domestic banks; and (ii) higher costs of servicing the debt, as financial liberalization eliminated compulsory below-market financing of the budget by the domestic banking system. As a result, and due to the growth of the debt, central government interest payments (including capitalized and accrued interest, on the basis of Bank of Greece data) 1/ rose from 6.7 percent of GDP in 1989 to 11.3 percent in 1992 (Chart 12).

CHART 12
CHART 12

GREECE Fiscal Developments

(In percent of GDP)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: Ministry of National Economy; and Bank of Greece.1/ Accrual basis.2/ Includes capitalized and accrued interest.3/ Staff estimate, based on data provided by the authorities. Breaks in the series in 1990. when military debt and short-term liabilities to the Bank of Greece were included in government debt; and in 1993, when the government issued bonds to replace foreign exchange valuation differences with the Bank of Greece.

The primary adjustment during 1990-92 was realized mostly through revenue increases: revenue rose by 6½ percentage points of GDP between 1989 and 1992 to over 29 percent. The bulk of the increase came from increased indirect tax revenue, while the share of direct tax revenue to GDP remained constant. Also, about ½ percentage point of the increase was due to increased EU transfers to the budget. Primary expenditures were also reduced by about 2 percentage points of GDP during the same period, mostly through reductions in the wage bill and pensions.

Progress in fiscal consolidation suffered a setback in 1993, when the fiscal situation deteriorated. The revenue shortfall relative to the ambitious budget target (over 3 percentage points of GDP) was particularly pronounced, and was the result of a combination of factors: (i) over optimistic estimates of the yield of certain measures; (ii) the relaxation of the tax collection effort in the runup to the October elections; (iii) the impact on revenue of the weaker-than-expected economic activity; (iv) shortfalls in VAT receipts due to the implementation of the Single Market act in January 1993; and (v) the cancellation of the planned privatization of OTE (the proceeds of which had been included in the 1993 budget). Overruns also appeared in total expenditures, particularly in transfers to public sector enterprises and in guarantees. As a result, instead of a primary surplus of some 4 percent of GDP the government was in broad primary balance, while the overall deficit rose to 13 percent of GDP.

2. Implementation of the central government budget in 1994

Against this background, the objective of the 1994 budget was to put the adjustment back on track. The budget targeted a primary adjustment of almost 2 percentage points of GDP to bring the primary surplus of the central government to Dr 431 billion (1.9 percent of GDP) and the overall deficit to Dr 3,144 billion (13.6 percent of GDP). 1/ The primary adjustment was to come entirely from higher revenues, which were expected to increase by almost 3 percentage points of GDP, while primary expenditures were targeted to increase by about 1 percentage point of GDP, due to higher investment.

In the event, the overall budget targets were met: the central government primary surplus is estimated at Dr 439 billion (2 percent of GDP) and the overall balance at Dr 3,134 billion (13.5 percent of GDP). Although the budgeted revenue increase failed to materialize (revenues went up by an estimated 1.3 percentage point of GDP relative to 1993, about half the targeted increase), the government cut investment spending to its 1993 level in relation to GDP, contained its current primary spending, and intensified its collection effort in the second half of the year, including through some minor new tax measures (notably increases in cigarette taxes and a tax on repos).

The revenue shortfall vis-à-vis the target was primarily due to indirect taxes. Direct tax revenue increased relative to 1993 (from 6.6 percent of GDP in 1993 to a record 7.6 percent of GDP), slightly above the 1994 budget target (Tables 17 and 18). This can be attributed to the success of the summary procedures for closing outstanding tax cases and clearing tax arrears introduced in early 1994 (the ambitious budget target for receipts from tax arrears was almost achieved), as well as the taxation of repos introduced in April 1994, which brought taxes on interest income over their original budget target. Indirect tax performance, however, was disappointing: VAT revenue actually declined slightly from its 1993 level to 7.4 percent of GDP, well below its budget target, while consumption taxes also fell slightly below their 1993 level. Among the consumption taxes, energy taxes performed particularly badly (which the authorities attribute to increased smuggling), more than offsetting the increased revenue from cigarette taxes. The poor indirect tax performance appears to also reflect a shrinkage of the tax base following the Single Market act in 1993 through increased tax evasion; although the effect of the act was expected to be temporary, indirect tax revenue from imported goods seems to have suffered a permanent decline. Nontax revenue of the Ordinary Budget was slightly above the budget target due to higher-than-expected receipts from Lotto, as well as greater-than-budgeted profit transfers from public enterprises. 1/ Finally, Investment Budget revenue was broadly on target, while EU transfers to SAGAP were below target (but these do not affect the deficit, as they are matched by equal expenditures). 2/

Table 17.

Greece: Ordinary Budget Revenue (accrual basis)

(In billions of drachmas)

article image
Source: Ministry of Finance.
Table 18.

Greece: Ordinary Budget Revenue (accrual basis)

(In percent of GDP)

article image
Source: Ministry of Finance.

Ordinary Budget expenditures were contained below budget (Table 19). The wage bill was on target (taking into account the part of the budgetary reserve set aside for personnel expenditure), partly due to central government employment declining in 1994 to 273.2 thousand (plus 25.1 thousand contractual staff) after a large (pre-electoral) increase in 1993 to 279.6 thousand (plus 27.4 thousand contractual staff). 3/ Grants were slightly over budget, due to higher-than-budgeted “other” grants. Interest payments (a usual area of budgetary overruns in the past), were this time below target (“corrected” for the amount of Dr 462 billion originally not included in the budget). Important budgetary savings, of the order of 0.8 percent of GDP, arose from the cuts in the Investment Budget. The bulk of the cuts affected projects in agriculture, industry, and forestry, while outlays on infrastructure and local authority projects were closer to the targets (Table 20). The total Investment Budget spending was 3.5 percent of GDP, the same level as in 1993. Agricultural subsidies through the SAGAP, at 3.3 percent of GDP, were below budget (this item is the single biggest category of EU transfers to Greece--Table 21). In terms of functional classification, developments in 1994 continued the trend of recent years, with the share of outlays for debt service rising at the expense of other categories, particularly health and education (Table 22).

Table 19.

Greece: Ordinary Budget Expenditures (accrual basis)

article image
Source: Ministry of Finance.

Does not include Dr. 462 billion (2 percent of GDP) of interest payments due on December 31, 1994 (but payable on January 2, 1995), which are, however, included in the provisional 1994 outcome.

Does not include capitalized and accrued interest.

Bank of Greece data.

Table 20.

Greece: Investment Budget Expenditure by Sector

(In billions of drachmas)

article image
Source: Ministry of Finance.
Table 21.

Greece: Budget Transfers from and to the European Union

(In billions of drachmas)

article image
Source: Ministry of Finance.
Table 22.

Greece: Central Government Expenditure, Functional Classification

(Accrual basis)

article image
Source: Ministry of Finance.

Including capitalized and accrued interest based on Bank of Greece data (Dr 128 billion in 1991, Dr 498 billion in 1992, Dr 360 billion in 1993, Dr 259.3 billion in 1994, and Dr 183.9 billion estimated for 1995).

3. Public entity and enterprise performance

The financial performance of public entities (the seven largest state social security funds) has improved somewhat after the pension reform of 1992, with their deficit down from 4-4½ percent of GDP in the late 1980s to under 3½ percent in 1993 and 1994 (Table 23). This, together with an increase in grants from the budget, has reduced their borrowing requirement (on an accrual basis) to less than half a percentage point of GDP (Dr 84 billion in 1994).

Table 23.

Greece: Public Entities Balance 1/

(In billions of drachmas: accrual basis)

article image
Source: Ministry of National Economy.

Covers seven public entities (including NAT-KAAN).

Excluding amortization payments.

The financial accounts of public entities on an accrual basis suffer from a number of statistical problems, due to improper accounting practices in the social security funds (some of which have not had audited balance sheets for a number of years). Cash data, collected by the Bank of Greece, show a sharply improved financial performance of public entities after 1992: in 1994, in particular, instead of the deficit (after transfers) of Dr 84 billion (0.4 percent of GDP) referred to above, cash data show a surplus of Dr 512 billion (2.2 percent of GDP). A task force was recently set up at the Ministry of National Economy to explain the large difference between accrual and cash data. This issue is of considerable importance, since social security funds are part of the general government, the definition used in the Maastricht convergence criteria. 1/

In contrast to public entities, the performance of public enterprises deteriorated somewhat in 1993 and 1994 (Table 24). After an improvement in the early 1990s, when their deficit fell to nearly 3 percent of GDP, the situation worsened in 1993 and 1994, when large overruns appeared in a number of expenditure categories (notably the wage bill) and the deficit increased to almost 4 percent of GDP. The largest loss-makers were the State Railways, the Athens. Bus Company (especially after its re-nationalization in 1994), Olympic Airways, and the Post Office: revenues of these companies were not sufficient to cover even their current spending (Table 25). For the public enterprise sector as a whole, increased state transfers kept the borrowing requirement low (under half a percentage point of GDP) in 1994.

Table 24.

Greece: Public Enterprise Balance 1/

(In billions of drachmas: accrual basis)

article image
Source: Ministry of National Economy.

Covers 45 major public enterprises.

Breakdown into components are estimates.

Surplus (+) or deficit (-).

Excluding amortization payments.

Table 25.

Greece: Operating Balance of Selected Public Enterprises

(In billions of drachmas: accrual basis)

article image
Source: Ministry of National Economy.

Including Urban Transport Corporation (since June 1994), Athens-Piraeus Trolley Buses, and Athens-Piraeus Electric Railways.

Thirty additional public enterprises.

4. PSBR financing, debt, and debt guarantees

Taking into account the central government, public entities, and public enterprises, the public sector borrowing requirement (PSBR) in 1994 on an accrual basis reached Dr 3,287 billion (14.2 percent of GDP), lower than its peak in 1989, but higher than in 1992 (11.7 percent) and 1993 (13.7 percent). On a cash basis, however, due to the large unexplained differences between accrual and cash accounts of public entities, the PSBR in 1994 declined to Dr 2,715 billion (11.7 percent of GDP), compared with 13.7 percent in 1993 (Table 26).

Public Sector Borrowing Requirement (PSBR), 1989-94

article image
Table 26.

Greece: Public Sector Balances

(In billions of drachmas)

article image
Sources: Ministry of Finance; Bank of Greece; and Fund staff estimates.

Does not include Dr 462 billion of interest payments due on December 31, 1994 (but payable on January 2, 1995), which are, however, included in the provisional outcome for 1994.

Includes capitalized and accrued interest based on Bank of Greece data amounting to Dr 128 billion in 1991, Dr 498 billion in 1992, Dr 360 billion in 1993, Dr 259 billion 1994, and Dr 184 billion estimated for 1995.

Covers seven social security funds (including NAT-KAAN).

Staff estimates; not comparable to official data on national accounts basis.

After the extraordinary increase in the share of external financing in 1993 to almost one-third of the total cash PSBR, which was due to the funding difficulties the budget faced late in that year, the composition of financing in 1994 returned to the pattern of the late 1980s and early 1990s. Almost 60 percent of the cash PSBR was financed by the domestic nonbank public, while the share of external financing was less than 10 percent (Table 27).

Table 27.

Greece: Financing of the PSBR

(In billions of drachmas)

article image
Source: Bank of Greece.

Including Treasury bills and bonds held by the Bank of Greece, as well as changes in the balance of the petroleum account through 1992.

A number of changes required in order to bring the definition of government debt in Greece in line with that used in the rest of the EU have affected the Greek government debt statistics in recent years. The most important among them were: (i) the inclusion of military debt in central government debt in 1993 (about 4 percent of GDP); the central government debt series have been correspondingly corrected back to 1990; (ii) the inclusion of the Treasury’s current account at the Bank of Greece, which was replaced with long-term government bonds in 1993, in government debt (about 4 percent of GDP); the debt series again have been correspondingly corrected back to 1990; (iii) the replacement of foreign exchange valuation losses on government liabilities held by the Bank of Greece with government bonds in 1993; this resulted in an increase in central government debt equivalent to about 10 percent of GDP in that year; and (iv) the classification of social security fund holdings of Treasury bills as intergovernmental debt in 1994, which reduced general government debt by about 3 percent of GDP in that year (leaving central government debt unchanged). Taking these changes into account, the central government debt at end-1994 was 121.8 percent of GDP, while the debt of the general government (Maastricht definition) was 114.1 percent of GDP, and that of the broader public sector 123.1 percent of GDP (Table 28).

Table 28.

Greece: Gross Public Sector Debt

(In billions of drachmas: end of period)

article image
Source: Ministry of Finance; and Bank of Greece.

Bank of Greece data. Interest capitalized during 1991 was added to the stock of debt at end-1992.

Replaced by Long-term bonds at and-1993.

Bonds issued in 1993 to cover valuation differences.

Break in the series; prior to 1994, Treasury bills held by the Bank of Greece on behalf of the social security funds were not registered as social security fund claims on the government.

Chart 13 presents the currency composition of central government debt in 1990 and 1994. In recent years, the authorities have attempted to limit net external borrowing (with the notable exception of 1993, as mentioned earlier). As a result, the share of external debt declined slightly from 24 percent of the total in 1990 to 22 percent in 1994. At the same time, however, the authorities increased recourse to foreign currency-linked borrowing (issuing bonds of 1-5 years’ maturity) which, until 1993, had been sold almost exclusively to domestic residents (although in 1994, large amounts of foreign currency-linked paper were also sold to foreigners). Thus, the share of foreign currency and foreign currency-linked debt increased from 32 percent of the total in 1990 to 37 percent in 1994.

CHART 13
CHART 13

GREECE Currency Composition of Central Government Debt

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Ministry of Finance.

The maturity structure of central government debt in 1990 and 1994 is shown in Chart 14. Overall, there has been a sharp reduction of the share of short-term Treasury bills (up to 12 months’ maturity) from 46 percent of the total in 1990 to 27 percent in 1994; also, short-term liabilities vis-à-vis the Bank of Greece (6 percent of the total in 1990) were replaced with long-term bonds at end-1993, when monetary financing was terminated. But most of this reduction was reflected in an increase of the share of floating rate bonds which, from the point of view of debt service, have the same characteristics as short-term debt. The share of medium- and long-term debt increased only slightly from 38 percent of the total in 1990 to 42 percent in 1994.

CHART 14
CHART 14

GREECE Maturity Structure of Central Government Dept

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Ministry of Finance.1/ Including consolidation bonds.2/ Including military debt.

In addition, at end-1994 there were Dr 2,133 billion (9.2 percent of GDP) of outstanding government-guaranteed debt, of which Dr 1,098 billion was external, and the rest was domestic debt. As regards the borrowers, the bulk of outstanding guarantees were for debt of public enterprises and local authorities; the government-guaranteed debt of private enterprises and individuals was only about Dr 245 billion at end-1994. The authorities have continued to issue new debt guarantees: in 1994, Dr 251 billion of new guarantees were issued, compared with Dr 273 billion in 1993. At the same time, significant amounts of guarantees are called every year as the original borrowers default on their payments, and the government either takes over these debts or pays them back; the process of settling guarantees that have been called often takes a long time, during which the debt is in dispute. In 1994, Dr 201 billion of guarantees were called, 1/ and another Dr 164 billion of guarantees were expected by the authorities to be called during 1995.

5. The 1995 budget

In line with the government’s revised convergence program, the 1995 budget targets a further primary adjustment of 1.4 percentage point of GDP to bring the primary surplus to Dr 859 billion (3.4 percent of GDP). This, and a decline in interest rates, would reduce the overall central government balance from an estimated 13.5 percent of GDP in 1994 to 10.6 percent of GDP (Dr 2,680 billion) in 1995, including capitalized and accrued interest. 1/

Broadly speaking, the 1995 budget puts the burden of the targeted primary adjustment again on revenue. The targeted increase in revenue is expected to more-than-offset the increase in investment spending, while primary current expenditures (excluding EU-financed agricultural subsidies) are targeted to remain unchanged in percent of GDP.

Ordinary Budget revenue is targeted to increase by some 18½ percent to reach 27.4 percent of GDP, compared with an estimated 25.3 percent in 1994. The bulk of this increase is expected to come from tax revenue. Personal income tax receipts are targeted to increase by over 30 percent and corporate income tax by 27 percent, as a result of the application of the new presumptive tax on professionals and the self-employed (see Box 4). 2/

The Presumptive Tax

The imposition of a presumptive tax, which was legislated in 1994 and took effect in 1995, was motivated by data of the General Accounting Office showing that the income declared to the tax authorities by a large number of professionals was implausibly low. For personal companies, the base of the tax is the national minimum wage for an employee with five years’ experience plus the imputed rent of the commercial property (6 percent of the imputed value of the property). This base is adjusted by a number of factors, mainly the location and type of the company (commercial, artisanal, etc.) and the number of years in operation. For professionals and self-employed people, the law specifies imputed income by category, which generally depends on the national minimum wage for an employee of similar qualifications, the imputed rent of the property (if applicable), and the years in business. Various adjustment factors are specified for each profession depending on the type (e.g., university degrees for economists, university positions for doctors, number of records for musicians, etc.). For farmers, the tax liability is determined by a committee established by the Ministers of Finance and Agriculture, based on the type of crop, size of holding, etc. Penalties are specified for farmers who fail to declare income under these provisions, including the termination of all subsidies or compensation for natural disasters, the prohibition to sell or transfer their land for five years, and restrictions to their access to EU assistance through the Common Agricultural Policy.

Property taxes and revenue from direct tax arrears are also projected to increase significantly (the latter due to the ongoing effects of the accelerated procedures for closing outstanding tax cases introduced in 1994). Ambitious increases are also targeted for indirect taxes, notably the VAT (which accounts for about half of all indirect tax revenue is expected to increase by almost 16 percent, as a stepping-up tax enforcement and the planned reorganization of Finance. Ordinary Budget nontax revenue is expected to as a result of higher EU transfers to the Ordinary Budget. Investment Budget and SAGAP revenues are projected to rise with higher EU transfers.

Data from the first five months of 1995 indicate that revenue collection has so far been successful: Ordinary Budget revenues grew by over 23½ percent over the first five months of 1994, compared with an annual target of 18½ percent. This overachievement of the budget target is largely explained by the fact that tax collection in the first few months of last year was very poor, as well as by the payments for settling the outstanding tax cases which started in April 1994 and are continuing in 1995. Nevertheless, even correcting for these factors, the underlying revenue performance is in line with the budget target. VAT performance, however, is a concern: data for the first four months indicate that at 14 percent, the increase in VAT revenue is below the budget target.

Ordinary Budget expenditures are targeted to decline (from an estimated 35.6 percent of GDP in 1994 to 34.7 percent in 1995) mostly as a result of lower interest outlays. Primary expenditures are budgeted to remain broadly constant in percent of GDP at their 1994 level. The interest payments projection is based on the assumption that the interest rate on the benchmark 12-month Treasury bill would average 15½ percent in 1995, some 4 percentage points below the 1994 average (at end-June, the 12-month Treasury bill rate was 15.75 percent). The wage bill (including the reserve fund for new hiring) is projected to increase by only 6.3 percent; excluding special personnel allowances, however, which are budgeted to return to their 1993 level (except for those paid to teachers and the new tax inspectors), the budgeted wage bill increase is 12½ percent. Civil servants’ pensions are budgeted to increase by 9.3 percent. A large increase in interest subsidies is budgeted (but the total amount, at Dr 20 billion, is small), due to higher subsidized loans to 1994 earthquake victims, as well as to business in Thrace (NE Greece). Other Ordinary Budget expenditure categories are projected to grow broadly in line with nominal GDP. Finally, as in 1994, Investment Budget spending is targeted again to increase substantially (by over 31 percent) to reach Dr 1,050 billion (4.1 percent of GDP, compared with an estimated 3.5 percent in 1994). The size of this increase reflects partly larger EU capital transfers, and partly higher investment spending financed by domestic means (EU capital transfers account for about 40 percent of Investment Budget spending).

VI. Inflation, the Budget Deficit and Debt Dynamics

It is a well known fact that inflation imparts an upward bias in conventionally-measured government deficits. Thus, it has been suggested that an inflation-adjusted measure (commonly referred to as the operational deficit) be used in countries where high inflation, a large stock of public debt, or a combination of both, is likely to lead to a sizeable bias in the conventional deficit. 1/ This Chapter discusses briefly the methodology of the inflation correction of the fiscal accounts, and presents an inflation-adjusted measure of Greece’s government deficit.

1. Inflation, the conventional deficit, and the operational deficit

In an inflationary environment, interest payments on government debt are overestimated, even if expressed as a percentage of GDP. The overestimation is due to the inflation premium embedded in the nominal interest rate, which is there to compensate for the inflation erosion of the real value of the debt. This means that conventionally-measured current expenditures include a partial amortization of the government debt. As a result, the conventional budget deficit overstates the actual impact of fiscal policy on government debt accumulation. It should be stressed that the inflationary bias arises only when the interest rate on government debt varies with inflation; when the public debt is indexed to inflation, or denominated in foreign currency, there is no inflation bias in the conventional deficit-to-GDP ratio.

To present these ideas more formally, assume for simplicity that nominal GDP (Y), government non-interest expenditures (G), and government revenues (T) grow at the rate of inflation (π) (i.e., they are fixed in real terms):

Y t = Y t − 1 ( 1 + π t ) ( 1 )
G t = G t − 1 ( 1 + π t ) ( 2 )
T t = T t − 1 ( 1 + π t ) ( 3 )

The conventionally measured deficit (Δ) and the stock of outstanding debt (D) are given by:

Δ t = ( G t − T t ) + i t D t − 1 ( 4 )
D t = D t − 1 + Δ t ( 5 )

where i is the nominal interest rate, given by i = (1 + r)(1 + π) - 1, and r is the real interest rate.

Rewriting equations (4) and (5) as ratios to GDP using (1)-(3) yields:

Δ t Y t = G t − 1 − T t − 1 Y t − 1 + [ r t + π t 1 + π t ] D t − 1 Y t − 1 ( 6 )
D t Y t = G t − 1 − T t − 1 Y t − 1 + D t − 1 ( 1 + r t ) Y t − 1 ( 7 )

Equation (6) shows that the conventional deficit as a proportion of GDP is affected by the level of inflation, while equation (7) shows that the evolution of the debt-to-GDP ratio is independent of (homogenous of degree zero in) the rate of inflation. Thus, the conventional deficit is not a reliable indicator of the effect of fiscal policy on debt accumulation, because it is affected by a variable that has no effect on debt dynamics. For the same reason, the conventional deficit should not be used for international comparisons if the inflation rates of the countries involved differ markedly.

The problem described above does not arise when domestic debt is indexed. To illustrate this point, it is necessary to make two changes to the basic equations (3) and (4). First, when domestic debt is indexed, interest payments are given by the product of a real interest rate and an inflation-adjusted stock of debt. Second, indexation implies that the stock of debt increases not only as a result of the deficit, but also directly due to inflation. The adjusted equations are:

Δ t = ( G t − T t ) + r t + ( 1 + π t ) D t − 1 ( 8 )
D t = D t − 1 ( 1 + π t ) + Δ t ( 9 )

Rewriting these equations as ratios to GDP using (1)-(3) yields:

Δ t Y t = G t − 1 − T t − 1 Y t − 1 + r t D t Y t − 1 ( 10 )
D t + 1 Y t = G t − 1 − T t − 1 Y t − 1 + D t ( 1 + r t ) Y t − 1 ( 11 )

The deficit is no longer a function of the level of inflation, and the debt dynamics equation is identical to that obtained for non-indexed debt. Thus, in this case, the deficit is a reliable indicator of the effects of fiscal policy on debt accumulation.

The same result is obtained when the public debt is denominated in foreign currency, because the effect of currency depreciation on interest payments operates through a nominal revaluation of the value of the debt, rather than through a higher nominal interest rate. If PPP holds, the level of nominal interest payments increases as the currency depreciates, but the level of nominal GDP increases proportionately and, as a result, the conventional deficit (as a percentage of GDP) is not distorted by inflation. 1/ Of course, if the real exchange rate changed, valuation gains or losses would add to or subtract from the debt.

In light of these results, it is clear that in order to eliminate the inflation bias in the conventional deficit-to-GDP ratio, one must subtract from it πtDt, the product of the inflation rate and the stock of floating interest debt. The resulting measure is usually referred to as the operational deficit.

2. Operational deficit in Greece and other European countries

This section presents estimates of Greece’s operational deficit for the period 1989-94, and compares these with those of other high-debt European countries. Operational deficits for the general government were constructed as the difference between the conventional deficit and an inflation adjustment term defined as the product of the average CPI inflation rate and the average stock of gross general government debt in domestic currency. 2/

The first of the tables below and Chart 15 show three indicators of fiscal performance: the conventional deficit, the operational deficit, and the primary deficit. Greece’s conventional deficit as a percentage of GDP improved by 2½ percentage points between 1989 and 1994; however, the operational balance suggests a somewhat stronger improvement (3 percentage points during the same period). The difference is due to a larger inflation component of interest payments in 1994 than in 1989, mainly as a result of a much larger stock of debt in 1994. This improvement, however, has not been continuous: while the primary deficit improved practically every year between 1989 and 1994 (reaching a cumulative improvement of almost 11 percentage points of GDP), the operational deficit worsened since 1993, when the real interest rate effectively paid on the government debt became positive. This rate had increased since 1990 as administratively set rates on government paper gradually began to approach their implicit market value, and reached 3½ percent in 1994.

CHART 15
CHART 15

GREECE General Government Deficits

(In percent of GDP)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Data provided by the authorities; and staff estimates.

The second of the tables that follow and Chart 16 compare Greece’s 1994 operational deficit with that of three other high-debt European countries: Belgium, Italy and Sweden. While Greece’s conventional deficit is clearly the largest of all, its operational deficit is smaller than Sweden’s and much closer to Italy’s than suggested by conventional deficits.

CHART 16
CHART 16

GREECE 1994 Inflation Adjusted Deficits: International Comparison

(In percent of GDP)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Data provided by the authorities; and staff estimates.

Conventional and Operational Deficits of the General Government

article image
Sources: Data provided by the authorities; and staff estimates.

1994 Inflation Adjusted Deficits: International Comparison

(In percent of GDP)

article image
Source: Data provided by the authorities, and staff estimates.

VII. The Medium-term Outlook for the Debt Dynamics

This Chapter discusses in more detail the staff’s medium-term scenario discussed with the authorities during the 1995 Article IV consultation and presented in SM/95/173 (see Table below).

Staff Medium-Term Scenario

(In percent of GDP. unless otherwise indicated)

article image

Accruals basis; different from data on national accounts basis for 1994 and 1995, but assumed to be equal thereafter.

Including capitalized and accrued interest in 1994-96 based on Bank of Greece data.

National accounts basis; including EU transfers to the Investment Budget.

The model used to generate this scenario is essentially a large-scale macroeconomic accounting framework with three main blocks: public finances and debt, the balance of payments, and the demand side of the national accounts. Three categories of variables are specified: environmental variables, which are given exogenously (and are consistent with the latest WEO projections); policy parameters, which are also controlled exogenously; and endogenous variables, generated on the basis of either behavioral equations or accounting identities.

The purpose of the scenario is to show the primary fiscal adjustment effort needed to achieve the overall deficit targets of the authorities’ convergence program. 1/ The convergence program was finalized in mid-1994, taking into account the international economic environment prevailing at the time, and making a number of key assumptions about domestic developments. Some of these (notably the interest rate assumption and the expected privatization receipts that would be used to reduce the debt) have been invalidated by developments in 1994-95. At the same time, the output recovery during this period has been stronger than anticipated at the time of the convergence program. These developments require a re-evaluation of the adjustment path of the program.

For comparability with the authorities’ convergence program targets, the scenario is expressed in terms of the general government. As discussed in Chapter V, data on the general government are problematic. This scenario is based on fiscal data (accrual basis), which are different from the national accounts data used in the convergence program. Thus, some differences exist between the staff estimates of government balances and those of the authorities for 1994 and 1995, but they are assumed to be zero thereafter.

It should be stressed that because Greece’s public debt is very high and predominantly short-term, any medium-term debt dynamics exercise is particularly sensitive to the interest rate assumption: a 1 percentage point increase in the world interest rate, if fully passed on to rates in Greece, would result in higher interest payments for the general government of 0.8 percent of GDP in the first year, growing to 1.1 percent of GDP in the third year. Given this sensitivity, as well as the statistical problems referred to above, this staff scenario should be seen as an illustrative exercise, rather than a forecast.

The growth outlook incorporated in the staff scenario envisages a relatively rapid recovery during 1995-97 (faster than that of the authorities’ convergence program), 2/ with growth reaching 3 percent in the medium term. The higher investment rates needed to achieve this can be attained through a combination of higher EU transfers under the Delors II package and higher private investment, spurred by the improved outlook for growth and profitability. Thus, the gross investment-to-GDP ratio is projected to increase from under 21 percent in 1994 to almost 28 percent by the end of the projection period. These investment rates are not historically high for Greece. Furthermore, structural reforms that reduce state intervention, combined with macroeconomic stability, may stimulate a more vigorous increase in investment and thus faster growth, but these effects are very hard to quantify.

Fiscal consolidation along the lines of the convergence program and continued tight monetary policy are expected to contribute to a gradual decline in inflation from an average of 10.9 percent in 1994 to under 3½ percent by 1999. This projection is also subject to considerable uncertainty since, aside from appropriate fiscal and monetary policies, gains in disinflation also depend on wage moderation.

The staff scenario assumes that domestic interest rates will continue declining, albeit much more slowly than assumed in the authorities’ convergence program. The benchmark 12-month Treasury bill rate is assumed to average 16.2 percent in 1995 (compared with a convergence program assumption of 14.1 percent and a 1995 budget assumption of 15.5 percent--see Chapter V). This is equivalent to a substantial reduction in both the expost real interest rate and the premium over German rates.

Staff scenario: Interest Rate Assumption

(in percent: period average)

article image

In order to generate the debt dynamics, the scenario makes specific assumptions about the composition of the government deficit financing. The scenario assumes that the policy of no new external borrowing is strictly applied during this period while, at the same time, the share of longer-maturity instruments is increasing. Thus, by the end of the period, only about 20 percent of the deficit is financed with short-term Treasury bills, and the rest is covered by drachma-denominated and foreign currency-linked bonds.

During the projection period, the debt stock is affected by operations other than deficit financing. On the one hand, privatization receipts of the order of Dr 100 billion are assumed to go towards reducing the debt during 1995-96. On the other hand, it is assumed that part of the outstanding government-guaranteed debt would need--as in the past--to be taken over by the government, thus increasing the debt stock (a total of about Dr 1.3 trillion over the projection period, or about half of the outstanding stock of government-guaranteed debt at end-1994; see Chapter V).

Under these assumptions, the scenario generates the primary surpluses needed to achieve the convergence program’s overall deficit targets (including capitalized and accrued interest) for 1996-99. It is shown that, at 7-7½ percent of GDP in the medium term, these are substantially higher than the surplus targeted in 1995, and higher by about 1½ percentage point of GDP than those included in the authorities’ convergence program. In combination with the economic recovery, which is now stronger than expected at the time of the preparation of the convergence program, these surpluses would allow a stabilization of the debt-to-GDP ratio already in. 1995, and a subsequent decline to just above 100 percent of GDP by 1999.

VIII. Monetary Policy and Interest Rates

1. Overview: the framework and instruments of monetary policy

In recent years, the Bank of Greece has been announcing target ranges for the growth of broad money (M3) and domestic credit while, at the same time, setting short-term interest rates at a level consistent with the official “hard drachma” policy (a rate of crawl of the drachma no higher than the targeted inflation differential between Greece and its EU partners). Instability in M3 velocity caused by financial sector liberalization and innovation in the late 1980s and 1990s led the Bank of Greece to also start announcing annual ranges for a broader liquidity aggregate (M4). 1/ The capital controls in place limited the instances of conflict between the exchange rate and the monetary targets; but when conflict did arise, the Bank increasingly chose to follow the exchange rate target, de-emphasizing the role of monetary targets.

Increasing signs of instability of M3--notably the massive tax-induced shift out of repos in 1994--and significant measurement problems affecting both M3 and M4, as well as the abolition of the last short-term capital controls in May 1994, have forced the Bank of Greece to focus on the exchange rate as its only intermediate target, at the expense of liquidity targeting. In its monetary policy statement for 1995, the targeted rate of crawl of the drachma vis-à-vis the ECU was explicitly announced for the first time (3 percent during the year), and although ranges were specified for M3 and M4 growth, these were not described as targets but “guidelines” (Table 29).

Table 29.

Greece: Monetary Program and Outturn 1/

(End of period)

article image
Source: Bank of Greece.

The definition of net domestic credit and credit to the public sector in the monetary program is different from that in the monetary survey; it includes borrowing by the public sector directly from abroad, but excludes capitalized interest. Also, for all credit aggregates the data do not reflect the exchange of government-guaranteed credit for government bonds.

Percentage changes in credit to the public sector and net domestic credit are calculated as the flows during the year excluding valuation adjustments over the stock of debt outstanding at the end of the previous year.

The main vehicle of intervention of the Bank of Greece in the money market has traditionally been the overnight interbank facility and, to a much lesser extent, swap transactions in the foreign exchange market. The official Lombard and rediscount facilities have been rarely used, although interest rates continue to be announced for them. 1/ The penalty overdraft facility, although also infrequently used (with the exception of the period of the speculative attack against the drachma in May 1994), has an important signaling effect. Interest rates for all these facilities are administratively set by the Bank. Finally, the reserve requirement on drachma deposits is 9 percent (Table 30).

Table 30.

Greece: Changes in Minimum Reserve Requirements and Investment Ratios for Commercial Banks

(In Percent of deposits)

article image
Source: Bank of Greece.

Applicable on drachma and foreign exchange deposits, surrendered to the Bank of Greece against drachmas.

Of which, as of June 1988 3 1/2 percentage points, and as of September 1991 4 1/2 percentage points, are non-interest-bearing.

Share of the increase in deposits that must be invested in Treasury bills.

In May 1993 this requirement was abolished.

Applicable only to drachma deposits.

As of November 1992, this requirement applied only to changes of deposits.

In July 1993 this requirement was abolished.

In November 1994, the Bank of Greece started intervening at the one-month end of the interbank market. This was made necessary by the need to sterilize large private capital inflows during this period (see Chapter IX). The size of the required sterilization operations, together with the attractiveness of the new facility for banks, quickly pushed the outstanding stock of 1-month interventions to over Dr 1 trillion by early 1995 (about 4 percent of GDP). 2/ The difficulty of simultaneously setting rates in the interbank market for two different maturities (overnight and one month), as well as the need to have a better way of gauging market sentiment, led the Bank to start weekly auctions of one-month deposits in the interbank market in May 1995, and of repos in June 1995.

2. Monetary and credit developments in 1994

Developments in monetary aggregates in 1994 reflected fundamental portfolio shifts resulting from the financial liberalization of recent years and the appearance of new financial products. These, along with other measurement problems described below, complicate the interpretation of the behavior of monetary aggregates in 1994.

Broad money (M3) growth during 1994 was 8.3 percent, within the range targeted by the Bank (8-11 percent) for the first time in some years (Tables 30 and 31). 3/ But this outturn reflects a move away from repos (following the extension of the 15 percent withholding tax on deposits to repos in April 1994) into assets not covered by M3, such as government securities (whose interest remains untaxed) and “synthetic swaps”. 1/ The substitution toward government securities affects less total drachma financial assets (M4), which grew at 13.4 percent during 1994 (about 12 percent in the year to May 1995). 2/ In real terms, M3 fell by 2.3 percent during 1994, while M4 grew at 2.4 percent, somewhat faster than estimated real GDP growth. Consequently, M3 velocity, which had already been unstable in recent years due to continuing financial liberalization, increased, while M4 velocity was more stable (Chart 17). Narrower monetary aggregates were also affected by these portfolio shifts: M1 grew by 25.4 percent in 1994, reflecting also some degree of substitution away from repos into sight deposits, as well as a reduction in the ratio of currency to sight deposits.

Table 31.

Greece: Monetary Survey 1/

(In billions of drachmas: end of period)

article image
Sources: Bank of Greece.

Revised data not comparable to previous years, due to a change in the reporting system. Data reflect the exchange of government-guaranteed credit for government bonds. Also, net credit to the central government in 1991-94 includes capitalized interest on government bonds held by commercial banks.

Includes securities and loans in foreign currency.

Excluding long-term loans in foreign currency.

The monetary aggregates are defined as follows: narrow money (Ml) is currency plus private sight deposits (excluding blocked deposits); broad money (M3) is M1 plus time and savings deposits, bank bonds and repurchase agreements; total drachma financial assets (M4) is Ma plus private sector holdings of T-bills and government bonds of maturity up to one year.

CHART 17
CHART 17

GREECE Money Growth, Velocity and Inflation

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: Bank of Greece.1/ End of period.2/ Ratio of nominal GDP to M3 and M4, respectively.

On the asset side, net foreign assets of the consolidated banking system increased significantly (Dr 1,600 billion) during 1994, primarily reflecting the accumulation of reserves at the Bank of Greece. Growth in net domestic credit, on the other hand, was 10.5 percent during 1994 (a reduction of 0.2 percent in real terms). Net domestic credit growth to the public sector fell significantly, from 22.8 percent during 1993 to 9.2 percent during 1994. This reflected both the decline in the cash public sector borrowing requirement (PSBR) and the relative success of net placements of government debt with the resident non-bank public (for a detailed discussion of the cash PSBR and its financing, see Chapter V). Net domestic credit to the private sector, on the other hand, grew by 12.5 percent (1.6 percent in real terms), more than the programmed 11 percent. This was mainly due to the rebound in economic activity, currently estimated at about ½ percentage point higher than expected at the time of the 1994 monetary program, and some natural recovery after very weak private credit growth during 1993. This increase in bank credit to the private sector was combined with a reduction in direct financing of private firms through the stock exchange (Table 33). As regards the sectoral distribution of private sector credit, above-average credit growth was registered in trade and housing, while the opposite was true in agriculture and manufacturing (Table 34).

Table 32.

Greece: Growth of Money and Credit Aggregates 1/

(In percent per annum; end of period)

article image
Sources: Bank of Greece; and Fund staff calculations.

Figures for 1991-94 include capitalized interest on government bonds held by commercial banks. Data also reflect the exchange of government-guaranteed credit for government bonds.

Excluding long-term loans to government in foreign currency.

Table 33.

Greece: Issues of Securities Through the Athens Stock Exchange

(In millions of drachmas)

article image
Source: Bank of Greece.
Table 34.

Greece: Distribution of Bank Credit to the Private Sector 1/

(End of period)

article image
Source: Bank of Greece.

Without taking into account the reduction in outstanding bank credit caused by the conversion of loans guaranteed by the government into government bonds. These conversions were: 1990 Dr 107.7 billion; 1991 Dr. 54.3 billion; 1992 Dr 185.0 billion; and 1993 Dr 492.1 billion.

The fact that significant increases in both net foreign and domestic assets were translated into a relatively small increase in liquidity (M3) is due to a very large fall in other items, net (Chart 18): the stock of other items reached Dr -3,875 billion at end-1994, a decline of Dr 1,801 billion. This change was unprecedented; in earlier years, net other items moved relatively little, mostly reflecting changes in bank capital. Unfortunately, lack of data at the time of the consultation prevented a complete analysis of the causes behind the decline in other items, net. But Bank officials speculated that the following were the most likely explanations: (i) the use of “synthetic swaps”, which banks cover through an inflow of foreign currency not recorded in the banks’ balance sheet until it is converted into drachmas (Bank staff thought that this factor alone could account for about half of the fall in net other items); (ii) sales of coupon payments from government paper as separate products, which are currently not included in liquid liabilities (money) and thus generate a decline in “other” net assets of banks; and (iii) sales of government paper by banks in the secondary market, which are also reflected in reductions in “other” net assets.

CHART 18
CHART 18

GREECE: Factors Affecting the Growth of Broad Money (M3)

(Flows in billions of drachmas)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Bank of Greece.1/ For 1991-1994 credit to public sector includes capitalized interest on government paper held by commercial banks.2/ Includes valuation changes on outstanding credit in foreign exchange.

3. Interest rates

After remaining broadly unchanged during the first four months of 1994, money market rates surged in May 1994 as the Bank of Greece successfully resisted a speculative attack against the drachma. Thereafter, money market rates declined rapidly to pre-crisis levels (about 19 percent by July). Since then, the Bank has been reducing interbank rates (including rates on the one-month facility, introduced in November) gradually: interbank rates fell by about 3 percentage points between July 1994 and April 1995 (Table 35). Recent (June 1995) auctions of one-month interbank deposits brought interest rates under 15½ percent. The overdraft, Lombard, and rediscount rates were also lowered in September and November 1994, and in March 1995 (Table 36).

Table 35.

Greece: Short-term Interest Rates

(In percent)

article image
Source: Bank of Greece.

Provisional data.

Table 36.

Greece: Official Interest Rates

(In percent)

article image
Source: Bank of Greece.

According to the size of the overdraft.

In addition, a penalty surcharge of 0.4 percent per day was imposed on bank overdrafts.

In addition, a penalty surcharge of 0.3 percent per day was imposed on bank overdrafts.

In addition, a penalty surcharge of 0.1 percent per day was imposed on bank overdrafts.

Commercial bank interest rates have traditionally been quite sticky in Greece, following the path of official rates with a substantial lag (Tables 35 and 37); 1/ there is also a considerable spread between lending and deposit rates (Chart 19). During 1994 and early 1995, however, bank rates have shown signs of increased flexibility. During 1994, bank lending rates fell by 1-2 percentage points, about the same as interbank rates. In early 1995, the decline in bank lending rates accelerated, reaching almost 3½ percentage points by April, with further declines announced by a number of major banks. Enterprise surveys indicate that the expectations of market participants about further interest rate reductions in 1995 are mixed (Box 5). Deposit rates followed a similar, but less steep pattern: they declined by 1-2 percentage points during 1994, but only marginally during the first quarter of this year. Recently (May-June), a number of major banks announced further declines in deposit rates. Overall, however, developments in 1994-95 have resulted in a small compression of commercial rate spreads.

Table 37.

Greece: Bank Interest Rates

(End of period; in percent per annum)

article image
Source: Bank of Greece.

The Lombard facility was introduced on June 16, 1993.

CHART 19
CHART 19

GREECE Interest Rates and Differentials

(In percent per annum)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

1/ Maximum rate charged by commercial banks on working capital loans to industry.2/ Treasury bill returns are tax free, whereas those for other assets shown are subject to a 15 percent witholding tax.3/ Rate offered by deposit money banks on three- to twelve-month drachma deposits.4/ On the basis of 12-month CPI inflation.5/ Ex post real rates on Greek Treasury bills minus ex post real rates on DM Eurocurrency deposits.

Interest Rate Expectations for 1995

(in percent of total reposes)

The opinion is split: almost half of industrial enterprise managers expect interest rates to decline in 1995 relative to 1994, but an equal percentage expects them to stay the same or increase (Source: ICAP Survey, 1995).

Interest rates on government paper are also administratively set. The benchmark 12-month Treasury bill rate (on which rates on floating Treasury bonds with longer maturities are pegged with a fixed margin) declined from 18 percent at end-1993 to 16 percent in April 1994, but was raised at the time of the foreign exchange crisis (Tables 38 and 39). 1/ Rates in the government paper market took longer than money market rates to return to pre-crisis levels, but the 12-month Treasury bill rate reached 18 percent again in November 1994. Since then, it has been lowered gradually and in June 1995 it reached 15.75 percent. Also in June 1995, the authorities held an auction of 3-, 5-, and 7-year bonds, for which the return of the first year was auctioned but the return for the remaining years was fixed. The auctions were successful, yielding first-year coupons of about 15½ percent.

Table 38.

Greece: Interest Rates on Government Paper

(End of period, in percent per annum)

article image
Source: Bank of Greece.
Table 39.

Greece: Foreign Exchange Government Bond Yields

(In percent)

article image
Source: Bank of Greece.

4. Changes in the regulatory environment of the banking system in 1994

Despite the financial liberalization of the late 1980s and 1990s, the Greek banking system is still dominated by a few large (mostly state-owned) banks. This oligopolistic structure is the main reason behind the stickiness of nominal interest rates and the persistence of the considerable interest rate spreads mentioned above. In recent years, however, including in 1994, there have been strong indications of increased competition in the banking system, as many new banks entered the market and the market share held by the large banks started declining.

The Bank of Greece has contributed to this outcome through a number of measures designed to liberalize the banking system and to harmonize the conditions under which banks compete. The only remaining institutions that are not currently operating under the uniform reserve requirement regime are the Building Societies, the Postal Bank, and the State Loan Fund (TΠΔ). In the first semester of 1994, the deposit base to which the reserve requirement is applied was broadened to include the deposits of public entities and public enterprises. In January 1994, consumer loans were liberalized when the credit limit per institution per client was raised to Dr 8 million. In August 1994, the requirement that term deposits in drachmas be held for a minimum of seven days was eliminated. And in early 1995, the redeposit requirement for foreign exchange deposits was lowered to 70 percent for most deposits.

These changes have already had some visible results. The specialized credit institutions have started to offer services previously dominated by commercial banks, 1/ while commercial banks have started to offer services previously dominated by the specialized credit institutions, thus increasing competition. 2/ Banks have become more sophisticated, investing in technology and offering new products, especially in foreign exchange transactions and risk management. A new electronic payments system (HERMES) is currently being implemented. Finally, Greek commercial banks have been very profitable in recent years, posting net income higher than the EU average, despite higher operating expenses; in 1994, bank profits in Greece reportedly rose by over 400 percent.

IX. Exchange Rate and the Balance of Payments

1. Exchange rate and international reserves

During 1994, the authorities continued to follow the “hard drachma” policy of keeping the rate of depreciation of the drachma vis-à-vis the ECU at or below the targeted CPI-inflation differential with the EU. By the end of 1994, the drachma had depreciated by 5.6 percent against the ECU, in line with the authorities’ objective. Achievement of this objective was threatened in May 1994 by a currency crisis, induced by expectations that the abolition of the last remaining short-term capital controls scheduled for July 1 would be accompanied by a step devaluation. The Bank of Greece resisted the speculative pressures raising its short-term interest rates to a peak of 180 percent, and abolishing short-term capital controls on May 16, ahead of the scheduled date. This response was successful, and by the end of May, pressures in the foreign exchange market had abated.

On average, the drachma depreciated by 6.7 percent against the ECU in 1994. The rate of depreciation of the nominal effective exchange rate was somewhat higher (7 percent), owing to the U.S. dollar’s weakening in international markets (Table 40 and Chart 20). 1/ However, as a result of the slowdown in domestic inflation in year-average terms, the CPI-based real effective exchange rate of the drachma stabilized in 1994, after increasing continuously since 1987. Real effective exchange rate indices based on unit labor costs (ULC) in the manufacturing sector, on the other hand, registered a small deterioration of cost competitiveness in 1994 compared to 1993, reversing a trend begun in 1991; this deterioration took place despite a reduction in real ULC in manufacturing in 1994. A detailed discussion and assessment of Greece’s external performance indicators and overall competitiveness position is provided in Chapter X.

Table 40.

Greece: Exchange Rates 1/

(Percentage changes)

article image
Sources: Bank of Greece; IMF, International Financial Statistics: and staff calculations.

Foreign currency per drachma; a negative sign denotes a depreciation.

Vis-a-vis May 1994.

Non-oil trade weighted vis-a-vis 15 competitor countries (1981-84 weights).

Based on OECD data. Unit Labor Costs in the business sector relative to a group of 18 industrial countries weighted by their share in Greece’s trade in 1992. The countries include all the EU countries, excluding Luxembourg, the United States, Japan, and the EFTA countries excluding Ireland.

April 1995 figure.

March 1995 figure.

CHART 20
CHART 20

GREECE Exchange Rates and Competitiveness

(1985=100)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: Data provided by the authorities; IMF, International Financial Statistics; and staff calculations.1/ Bank of Greece index.

In 1995, the authorities announced for the first time the targeted rate of depreciation vis-à-vis the ECU: 3 percent by year-end. During the period of market turbulence in early 1995, triggered by the collapse of the Mexican peso and, more recently, by the collapse of the U.S. dollar. the drachma has shown remarkable stability and resilience, despite realignments of the Spanish peseta and the Portuguese escudo and relatively strong depreciations of the Italian lira and the pound sterling. In addition, enterprise survey data show that the current exchange rate policy is credible (Box 6). The drachma depreciated by 1.4 percent against the ECU between December 1994 and May 1995, and 0.3 percent in nominal effective terms between December 1994 and March 1995.

International reserves recorded a substantial increase in 1994, with gross official reserves reaching a historical peak of US$15.3 billion by year-end (Table 41); they increased further to US$16.5 billion by end-May 1995. This outcome is to a great extent the effect of a surge in capital inflows (see below) that resulted from the restoration of confidence in the “hard drachma” policy after the successful handling of the May 1994 currency crisis, and to high domestic interest rates.

Table 41.

Greece: Official Reserves

(In millions of U.S. dollars; end of period)

article image
Sources: Bank of Greece; and IMF, International Financial Statistics.

National valuation.

Gold and foreign exchange holdings exclude the deposits made with the European Monetary Cooperation Fund (EMCF) of a portion of the gold and gross U.S. dollar holdings; the holdings of ECU issued by the EMCF against these deposits are included in foreign exchange holdings in this table.

Exchange Rate Expectations for 1995

(In percent of total responses)

The majority (75 percent) of industrial enterprise managers expect the policy of “hard drachma” to continue in 1995. Only one-tenth expect a rapid crawl of the drachma, and an even smaller percentage foresee a step devaluation (Source: ICAP Survey, 1995).

2. Balance of payments and external debt

This section discusses balance of payments developments based primarily on bank settlements data, which provide a more detailed breakdown of the current account than the customs-based national accounts data; references to customs data in the discussion of the current account are explicitly identified. Unfortunately, the reliability of settlements data has deteriorated in recent years as a result of various developments. In particular: (i) as a result of capital liberalization, export revenues no longer have to be repatriated through the domestic banking sector; (ii) foreign trade in drachmas with Balkan countries, not captured by settlements data, has been growing; and, (iii) some export earnings are recorded as “other invisible receipts” by banks. A detailed description of the differences between settlements data and customs data is provided in Appendix IV to the Recent Economic Developments document for the 1994 Article IV Consultation (SM/94/173).

a. Current account transactions

During 1994, the current account deficit continued to narrow, reaching 0.1 percent of GDP--a historically low level (Table 42 and Chart 21). The improvement from the 0.8 percent deficit recorded in 1993 was due to an increase in the invisibles balance, especially tourism (Tables 43 and 44), as EU transfers and the trade deficit remained unchanged at 4.5 percent and 14 percent of GDP, respectively.

Table 42.

Greece: Balance of Payments

(In millions of U.S. dollars: on a settlement basis)

article image
Sources: Bank of Greece, Monthly Statistical Bulletin: data provided by the authorities; and IFS.

Private and official capital, excluding errors and omissions.

Includes direct investment and enterprise borrowing abroad.

Includes official suppliers’ credits.

Borrowing by the Hellenic Industrial Development Bank, the Agricultural Bank of Greece, and the National Mortgage Bank of Greece.

CHART 21
CHART 21

GREECE Balance of Payments and Reserves

(Banking settlement basis)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: Bank of Greece; and IMF, International Financial Statistics.1/ Settlements data underestimate imports and --particularly--exports in the 1990s.
Table 43.

Greece: External Services and Transfers

(In millions of U.S. dollars)

article image
Source: Bank of Greece.

Receipts reflect net EU transfers.

Table 44.

Greece: External Current Account Deficit and Financing

(In percent of GDP. settlement basis)

article image
Sources: Bank of Greece; and IMF, International Financial Statistics.

In the trade account, both imports and exports rose in 1994 by 6 percent and 4 percent, respectively, following declines of 11.5 percent and 16.2 percent in 1993. According to estimates based on customs data, import and export volumes in 1994 rose by 3.7 percent and 3 percent, respectively, compared to a 1.3 percent increase in import volume and a 0.5 percent decline in export volume in 1993. The slower growth of exports in 1994 relative to 1993 was accompanied by a small decline in Greece’s export market share.

The increase in the invisibles surplus by 13 percent in 1994 was primarily due to stronger travel receipts, which grew by 17 percent from 3.7 percent of GDP in 1993 to 4.1 percent in 1994. Tourist arrivals expanded by 12 percent in 1994 mainly as a result of two factors: (i) the economic recovery in Europe, where more than 90 percent of tourists in Greece originate; and (ii) the diversion of tourism away from the former Yugoslavia and Turkey, due to security problems in these countries. Moreover, settlements data on tourism receipts underestimate actual tourism activity because some payments of tour operators are classified in other categories of the services account “convertible drachma accounts”) and also because tourism-related transactions that do not involve foreign exchange are not captured at all in the settlements data. EU transfers remained flat in 1994 relative to 1993, mostly due to delays in the execution of large-scale infrastructure projects.

Although the EU still remains by far Greece’s largest single trading partner, in 1994 the geographical distribution of exports shifted somewhat toward developing countries (as in 1993), while the share of imports from industrial countries increased (Table 47). The share of exports to industrial countries fell by 3 percentage points to about 67 percent of total Greek exports, while and the share of exports to nonoil-producing developing countries rose by 4 percentage points. Greece increased further its exports to Eastern European countries, with exports to Albania and Romania growing by 211 percent and 80 percent respectively. 1/ At the same time, the share of total Greek imports from nonoil-producing developing countries to total Greek imports fell by 1 percentage point, while imports from EU countries increased by over 6 percentage points. The increase in the share of imports from EU countries (mainly capital goods) probably reflects the strong growth in domestic investment in non-dwelling equipment in 1994.

Table 45.

Greece: Current Account of the Balance of Payments 1/

(In millions of U.S. dollars)

article image
Source: Ministry of National Economy.

National accounts presentation. Converted to U.S. dollars using the annual average exchange rate.

PIP: Public Investment Program.

Table 46.

Greece: Selected Indicators for Trading Partners 1/

(Annual changes, in percent)

article image
Source: International Monetary Fund.

Except for non-fuel commodity prices (see footnote 7 below), these composites are averages of percentage changes of data for each trading partner (as specified in footnotes 3 and 4 below) weighted by their share in exports or imports, as appropriate, of Greece.

Weights are proportional to 1992 exports of Greece to partner countries as specified in footnotes 3 and 4 below.

Based on data for partner countries that together account for at least 90 percent of exports or imports, as appropriate, of Greece.

Based on data for industrial partner countries only.

Weights are proportional to 1992 imports of Greece from partner countries as specified in footnotes 3 and 4 above.

That is, weighted averages of percentage changes in indices expressed in national currencies of industrial partner countries.

Based on averages of world market prices for component non-fuel commodities weighted by the 1979-81 composition of commodity trade (exports and imports) of Greece.

Table 47.

Greece: Geographical Distribution of Trade

article image
Source: IMF, Direction of Trade Statistics.

b. Capital account transactions and external debt

The capital account surplus increased by US$2.5 billion in 1994 to US$6.9 billion, mainly on account of large private capital flows, which expanded by 134 percent relative to 1993 to US$3.8 billion (Chart 22); net official capital flows, on the other hand, remained flat at about US$3 billion (Table 48). As a result, the “basic balance” (current account plus net private capital flows) improved substantially in 1994 due to the combination of a shrinking current account deficit and growing net capital inflows.

CHART 22
CHART 22

GREECE Private Capital Flows and Overall Balance

(In US$ millions)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Bank of Greece; and staff estimates.
Table 48.

Greece: Capital Account

(In millions of U.S. dollars)

article image
Source: Bank of Greece.

Includes some debt-creating capital flows in the form of enterprise borrowing abroad.

Long-term.

Borrowing by the Helenic Industrial Development Bank, the Agricultural Bank of Greece, and the National Bank of Greece.

Net private capital inflows increased markedly after June 1994, once the foreign exchange crisis was over, the credibility of the authorities’ exchange rate policy enhanced, and investor confidence restored. Unfortunately, statistical problems in the Greek balance of payments do not allow a detailed analysis of the inflows. Nevertheless, some information can be gleaned from the available data. A major category of the capital account affected by the inflows was “entrepreneurial capital”, which registered inflows of US$1,860 million between June and December. Of this, a large part appears to have been due to foreign borrowing by domestic firms seeking to lower interest costs; this was also spurred by the liberalization of short-term capital controls on May 16, 1994. 1/ Portfolio investment by foreigners in the Athens stock exchange (an estimated US$100 million in 1994 as a whole), as well as small inflows associated with the sales of state casino licenses (US$33 million in 1994) are also included in “entrepreneurial capital”.

Another major contributor to the surge in net private capital inflows was “other capital,” which includes mainly short-term, largely speculative capital movements. This category was the most affected by the May 1994 currency crisis: in May alone, there was an outflow of US$1.3 billion, but immediately after the end of the period of the crisis, this capital began to flow back, and during June-December this category registered inflows of US$1.5 billion.

Regarding the smaller items in the capital account, short-term suppliers’ credit, together with official suppliers’ credit, increased by US$701 million in 1994, mostly due to the rise in the commodity import bill, whereas the change in foreign exchange deposits by individuals--which reflect mostly the movements of emigrants’ foreign exchange accounts in Greek financial institutions--was positive during the current crisis because of the decrease in withdrawals from these accounts for conversion into drachmas,, and turned to small negative number afterwards.

Total public and publicly-guaranteed external debt, including suppliers’ credits, rose by US$5.4 billion to the equivalent of 34 percent of GDP by year-end (Table 49). 1/ The increase was driven mainly by medium- and long-term borrowing by the central government (US$8.6 billion). The debt service ratio increased by 4 percentage points to 35 percent, due exclusively to higher amortization payments, as interest payments remained unchanged (Table 50). Half of the US$1.4 billion increase in amortization payments is due to higher amortization of central government debt, with the Bank of Greece and the private sector recording also substantial increases in amortization payments.

Table 49.

Greece: External Debt 1/

(In millions of U.S. dollars: end of period)

article image
Sources: Bank of Greece.

Public and publicly-guaranteed debt and private nonguaranteed suppliers’ credit. As of end-1988, the authorities stopped publishing private external debt data; this amounted to US$2,840 million at the end of 1988.

Disbursed and outstanding; excluding military debt.

Short-term, as well as medium- and long-term; a breakdown by original maturity is only available for total (publicly-guaranteed and private) suppliers’ credits.

Table 50.

Greece: External Debt Service

(In millions of U.S. dollars)

article image
Source: Bank of Greece.

Medium- and long-term only. Includes both interest and amortization payments.

Debt service (total: A + B + C) in percent of current account receipts.

3. Changes in the regulatory environment

Two important changes in the framework regulating Greek international transactions took place in 1994: the liberalization of capital movements and the approval of the Uruguay Round.

The gradual liberalization of capital movements had started a number of years ago, and the last remaining short-term controls were liberalized on May 16, 1994. Only a few minor restrictions remain in foreign investments in border regions and in maritime transport, acquisition of mining rights, and foreign participation in enterprises engaged in radio and television broadcasting or air transport. In line with EU regulations, public procurement has also been opened to foreign competition.

The Greek Parliament has ratified the Uruguay Round. Adoption of the Uruguay Round norms, which require gradual tariff reductions (with half of the decrease to take place by 2004) may have adverse consequences for Greece in the textile and agriculture sectors. In textiles, Greece would face increased competition with imports from outside the EU, both domestically and in the markets of other EU members. However, some of the funds under the Community Support Framework would be allocated to improve the competitiveness of these sectors. The agricultural sector, which has already been affected by the gradual decline in subsidies provided under the CAP and the phasing out of subsidized credit, will be further affected by trade liberalization, especially in the area of exports of Mediterranean products, such as olive oil.

X. Competitiveness and External Performance

This Chapter assesses the international competitiveness of the Greek economy on the basis of two classes of indicators: direct indicators, including the standard measures of relative costs or prices between Greece and a weighted average of its trading partners (real effective exchange rates, or REERs); and other economic variables which are indicative of a country’s international competitiveness.

The need to extend the analysis beyond the various REERs arises from the shortcomings associated with the use of these indices as indicators of competitiveness. First, when traded products are imperfect substitutes and there are differences in growth rates and income elasticities across countries, the REERs may fluctuate independently of changes in competitiveness. Second, when productivity is higher (lower) in the tradable sector than in the nontradable sector, the observed real appreciation (depreciation) in price-based REERs does not necessarily reflect a loss of competitiveness (Balassa 1964). Third, different REERs often diverge significantly, complicating the assessment of competitiveness. More broadly, the fact that movements of REERs are often uncorrelated with changes in countries’ external positions, e.g., in the case of Japan (IMF 1994), suggests that REERs are imperfect indicators of overall competitiveness. For this reason, this Chapter examines some indirect indicators of competitiveness as well, notably the size and composition of the external current account, and various measures of export performance.

The Chapter is structured as follows: section 1 discusses in some detail the pros and cons of direct indicators; section 2 assesses Greece’s competitiveness using direct indicators computed for the period 1970-94; section 3 examines Greece’s current account and analyzes in more detail Greece’s export performance, by applying the constant-market-shares method; and the last section draws the main conclusions.

1. An overview of direct indicators of competitiveness

This section reviews the characteristics of the various direct indicators of competitiveness grouped according to the main economic variable on which they focus. Price-based indicators include the price-based REERs and the internal real exchange rate, and are constructed on the basis of the CPI, the WPI, or the GDP deflator. Cost-based indicators (cost-based REERs) utilize unit labor costs for the economy as a whole (ULC-based) or--more often--for the manufacturing sector only (ULCM-based). 1/ Profit-based indicators include absolute and relative profitability. Finally, export-based indicators include the export unit price (EUP)-based REER, export profitability, and the export market share.

a. Price-based indicators

The CPI-based REER has several advantages, for which it is the most frequently-used indicator: the CPI is based on a basket of goods that is fairly comparable across countries, is readily available, and is published frequently. At the same time, however, the CPI is subject to the influence of price controls, indirect taxes, and other distortions.

The WPI-based REER is used as frequently as the CPI-based REER. Unlike the CPI-based index, the WPI-based REER is not affected by changes in indirect taxes, distribution costs, or the price of services directly sold to consumers; however, it tends to underestimate changes in competitiveness, because the prices of many of the relatively homogeneous products in the WPI (raw materials and intermediate goods) do not differ much across countries. Also, the WPI is subject to problems of double-counting, since the prices of some intermediate goods are also directly reflected in the wholesale prices of final goods.

The GDP deflator-based REER reflects more closely a country’s domestic price competitiveness because imports are subtracted from expenditure to calculate GDP (Kenen and Pack 1980). When the country experiences high imported inflation, the GDP deflator-based REER tends to fluctuate less than the CPI- and WPI-based REERs. Unfortunately, GDP deflators are usually available with long lags.

The CPI-, WPI-, and GDP deflator-based REERs are often interpreted as the ratio of the relative prices of nontradable goods to tradable goods in a country; 1/ thus, an increase in these indicators would imply either a loss of competitiveness in the country’s tradable sector or a greater incentive to reallocate resources to the nontradable goods sector, or both (Feldman 1994). But this interpretation assumes that traded goods across countries are perfect substitutes, an unrealistic assumption.

The ratio of tradable to nontradable goods in a country (the internal real exchange rate) can be constructed directly by using the prices of these goods. An increase in the internal real exchange rate may be interpreted as a deterioration in competitiveness to the extent that it reflects a domestic resource pull toward the nontradable sector, which, in turn, may lead to a deterioration in the current account balance. However, the relative price of nontradables could also increase when there are no changes in competitiveness. For example, faster productivity growth in the tradable sector than in the nontradable sector may lead to a large appreciation even though the trade balance might at the same time be improving. In addition, a reallocation of resources to the nontradable goods sector is often appropriate to ensure a balanced availability of tradable and nontradable goods, and thus the real appreciation does not necessarily have adverse effects on the country (Lipschitz and McDonald 1991).

b. Cost-based indicators

The ULCM-based REER reflects the relative labor costs per unit of output of a country’s manufacturing sector relative to that of its trading partners. It is the most common cost-based indicator because the data for this sector are usually more reliable and easily available than the data for other sectors. However, the ULCM-based REER covers only direct labor costs, and only in one sector, which, for many countries, may be the source a relatively small share of its exports and imports (Feldman 1994). The ULC- based REER for the economy as a whole, if available, solves the second problem, but still does not cover important cost components, notably intermediate goods, the cost of capital, etc. Furthermore, a shift in the capital/output ratio may affect the ULC-based REER even if labor costs remain constant (IMF 1994).

c. Profit-based indicators

Absolute profitability is estimated by calculating the ratio of a country’s WPI to its ULC, while the ratio of relative prices to relative ULCs measures relative profitability. A decline in this indicator occurs when the increase of relative ULCs vis-à-vis the country’s trading partners is faster than that of relative prices, suggesting that relative profit margins are been squeezed, and perhaps foreshadowing a decline in output (Lipschitz and McDonald 1991). These indicators are not a good guide to developments in rates of return on capital, however, if there are significant differences in technology. For example, they do not reflect the increase in the return on capital that may arise as capital per unit of output falls more than profits per unit of output (Lipschitz and McDonald 1991).

d. Export unit price-based indicators

The EUP-based REER reflects a country’s price competitiveness in its export sector. On the face of it, this appears to be the most relevant indicator of price competitiveness. However, it has some serious shortcomings: it does not reflect competitiveness when suppliers adopt a pricing-to-market strategy by squeezing their profits in order to maintain their market shares; it fails to capture the prices of goods that are no longer exported because their domestic production costs exceed the prices at which they would be internationally competitive (IMF 1984); and it does not necessarily reflect competitiveness when a exogenous shift in demand preferences changes export prices.

Comparing the CPI- and EUP-based REERs, Marsh and Tokarick (1994) conclude that the former is preferable because it provides information on the prices of exports and imports rather than on just one component of the trade balance. The CPI-based REER also includes the price of exportable goods purchased by domestic consumers, while the EUP-based REER uses only the prices of exported goods. On the other hand, the EUP-based REER includes more information than the ULC- or ULCM-based REERs, such as the costs of other inputs in the traded goods sector; but the latter include information on ULC in the importable sector, while the former does not (Marsh and Tokarick 1994).

2. Assessing Greece’s competitiveness: direct indicators

This section assesses developments in Greece’s competitiveness by using direct indicators. Limited by data availability, series of REER indices were constructed on the basis of the CPI, WPI, GDP deflator, EUP, ULC, and ULCM, and various price ratios of services to manufactured products were calculated as a proxies for the internal real exchange rates on an annual basis from 1970 to 1994.

Chart 23 shows price-based REERs constructed on the basis on the CPI, the WPI, and the GDP deflator. These indicators diverged markedly during the 1970s, but began to follow similar patterns in the early 1980s. All three indicators show an unambiguous downward trend from 1983 to 1986 and an upward trend since the late 1980s. Most indicators, however, stabilized in 1994, as a result of the containment of domestic inflation.

CHART 23
CHART 23

GREECE Price-based Real Effective Exchange Rates 1/

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Sources: IMF, International Financial Statistics; IMF, Direction of Trade Statistics; and data provided by the authorities.1/ The main CPI-, WPI-, and GDP deflator-based indices are calculated with 1992 total trade weights from DTS for 18 countries. The BoG CPI- and WPI-based Indices use average total trade weights for 1981-84 for 15 countries. The IFS CPI-based Index uses average total trade weights for 1980-82 for 18 countries.

From the mid-1970s to the early 1980s, the CPI-based REER showed a real appreciation due to an increase in imported inflation caused by the two oil crises. The imported inflation affected the Greek economy more severely than other European economies because of the high degree of import penetration in Greece. A cumulative CPI-based real appreciation of the order of 35 percent took place between 1975 and 1982, reflecting also the effect of nontradable price inflation. The WPI-based REER followed a similar pattern, but the cumulative WPI-based real appreciation between 1975 and 1982 was only 14 percent. On the other hand, the GDP deflator-based REER was on a declining trend during the same period, mainly because the effects of imported inflation are filtered out when imports are subtracted from expenditures to calculate GDP.

Between 1983 and 1986, price-based REERs declined owing to a combination of factors, notably: two devaluations of the drachma (in 1983 and 1985), each by 15 percent; a containment of costs and prices due to controls introduced after the second devaluation in the context of the 1985-87 stabilization program; and a deceleration of inflation in housing and rents as a result of a regime of extensive controls.

Since the late 1980s, all three price-based REERs have again shown an upward trend. Following a decline of 7 percent in 1986 owing to the stabilization program, the CPI-based REER appreciated by 18 percent between 1986 and 1994 mainly due to the introduction of the value added tax in January 1987. The WPI- and the GDP deflator-based REERs also began to increase moderately in 1987, after the relaxation of policies following the 1985-87 stabilization program, but by 1994, both were below their average levels during the period 1970-94 as a whole.

Due to the lack of data on tradable and non-tradable prices, three proxies of the internal real exchange rate were constructed (Chart 24): (i) the consumer price of total services relative to that of manufactured goods (top panel); (ii) the consumer price of a subgroup of services taken to proxy nontradable services relative that of manufactured goods (top and bottom panels); 1/ and (iii) the ratio of the value-added deflator in the services sector to that in manufacturing (middle panel).

CHART 24
CHART 24

GREECE Internal Real Exchange Rates

(Relative price of nontradable to tradable goods)

(1980=100)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Bank of Greece; Ministry of National Economy; IMF, International Financial Statistics; and OECD, analytical data base.1/ Old national accounts data; available only through 1992.

All three indicators show an upward trend since the late 1980s, as is the case with price-based REERs. This has been a common phenomenon in other European countries. The middle panel compares the third indicator with estimates constructed for Italy and Spain: in all countries the indicators show an increase in the relative price of services; indeed, the extent of the real appreciation in Greece is considerably smaller than in the other two countries. The appreciation of the internal real exchange rate in several European countries has been attributed to labor market regulation, which has determined uniform rates of wage increases economy-wide, combined with faster productivity growth in the tradable goods sector.

Chart 25 shows the behavior of three cost-based indicators based on ULCs in the manufacturing sector (normalized and unadjusted) and the business sector (defined as the nonagricultural sector excluding government, state enterprises, and banks). All three follow similar paths to those exhibited by the CPI-based REER through the late 1980s. After that, however, they follow a very different path: an increase between 1987 and 1990 and a decline between 1990 and 1993, with a reversal in 1994.

CHART 25
CHART 25

GREECE Cost-based Real Effective. Exchange Rates 1/

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: IMF, International Financial Statistics; IMF, Research Department; IMF, Direction of Trade Statistics; and data provided by the authorities.1/ REERs based on NULC in manufacturing (IMF RES data) and ULC in the business sector (OECD data) calculated with 1992 total trade weights from DTS for 18 countries. The BoG ULC-based REER does not use normalized ULCs and uses average total trade weights for 1981-84 for 15 countries.

Between 1970 and 1972, the ULC-based REER declined by 29 percent owing to productivity increases and the suppression of trade unionism during the dictatorship. From 1974 to the mid-1980s, this was reversed, as labor costs ballooned; the wide fluctuations of cost-based REERs during these years reflect frequent shifts in Greece’s incomes policy, in particular the introduction of wage indexation in 1982.

The 1985-87 stabilization program led to a substantial improvement in Greece’s cost competitiveness: under the program, the indexation system was modified in order to link wage increases to targeted, rather than actual, inflation rates and to exclude the impact of import prices. However, after the stabilization program was abandoned in 1987, relative ULCs increased again, reflecting the generous pre-electoral wage increases granted during that period.

Since 1991, Greece has again gained ground in cost competitiveness through both a tight incomes policy and productivity gains. The wage indexation system was abolished, personnel in “ailing” firms was laid-off, and the anti-inflationary exchange rate policy also helped to exert downward pressures on costs. The slight real appreciation in 1994 took place since Greece’s trading partner countries, particularly France and Germany, outperformed Greece in containing labor cost increases. Nevertheless, at the end of the period, the cost-based indicators for Greece were at or below their average value during 1970-94.

Chart 26 shows various indicators of profitability, which combine information on WPI- and ULC-based indices. The top panel shows indices of absolute profitability in the business and the manufacturing sectors. After a large decline in profitability in the late 1970s, the trend has been slightly downwards. But relative profitability has been maintained broadly unchanged throughout the 1980s, despite wide cyclical fluctuations. Relative profitability clearly improved in the 1990s, although the increase in ULCs checked this trend in 1994.

CHART 26
CHART 26

GREECE Profitability

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: IMF; International Financial Statistics; IMF, Direction of Trade Statistics; IMF, Research Department; and data provided by the authorities.1/ Based on WPI and ULC in the business sector (OECD data)2/ Based on WPI and NULC in the manufacturing sector (RES data).3/ Calculations based on 1992 total trade weights from DTS, based on relative WPI and relative ULC in the business sector.4/ Calculations based on 1992 total trade weights from DTS, based on relative WPI and relative ULC in the manufacturing sector.5/ BoG calculations based on average total weights for 1981-84, based on relative WPI and relative ULC In the manufacturing

Chart 27 reviews export-based indicators. The top panel shows the EUP-based REER. It has followed a downward path between 1970 and 1987, registered an appreciation between 1987-89, and has since returned to a declining trend, that was reversed only in 1993.

CHART 27
CHART 27

GREECE Export Performance

(1970=100)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: OECD, OECD Economic Outlook; OECD, Analytical Database; IMF; International Financial Statistics; IMF, Direction of Trade Statistics.1/ EUP-based REER calculated with 1992 export weights from DTS.2/ EUP divided by unit labor costs (ULC in the business sector from the OECD, and NULC in manufacturing from RES. respectively).3/ OECD data, spliced with own calculations based on export volumes in IFS and market growth in GEE before 1978.

While a declining EUP-based REER suggests a gain in Greece’s export competitiveness, the indicator should be used with caution. Greece’s major exported goods are primary and labor-intensive manufactured products (e.g., food, beverages, tobacco, and textiles), while its major trading countries, which are highly-industrialized OECD countries, export manufactured and high technology-intensive products. This structural difference suggests that the EUP-based REER suffers from a composition effect. A real depreciation could simply reflect a decline in the prices of labor-intensive products relative to those of capital- or technology-intensive products in the face of tougher competition for the former from Greece’s neighboring countries.

The middle panel of Chart 27 shows absolute profitability in the export sector, which combines information from EUPs and ULCs in the business and manufacturing sectors. In both cases, absolute profits in the export sector have declined continuously since 1970, with the decline being stronger in the business sector. The downward trend reflects the squeeze in profit margins of Greek exporters that resulted from increased competition in export markets and the “hard-drachma” policy.

The bottom panel of Chart 27 shows an index of market share for Greek exports. Three episodes of rapid market share gains can be observed: 1970-76, 1982-87, and 1988-93. Overall, although Greece’s export market share has been maintained during the period, it has increased only modestly relative to Portugal’s remarkable increases throughout the 1980s and Spain’s gains in the 1990s (Chart 28).

CHART 28
CHART 28

GREECE Comparison of Export Performance

(1980=100)

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: OECD, OECD World Economic Outlook.

3. Assessing Greece’s competitiveness: indirect indicators

This section reviews two indirect indicators of international competitiveness, used to supplement the analysis of direct indicators: the current account and the performance of Greece’s exports.

a. The current account of the balance of payments 1/

Charts 29 and 30 show various indicators of the current account and its components. During the 1970s and early 1980s, the current account deficit fluctuated at about 4½ of GDP. Following a substantial deterioration in 1985, the current account has improved constantly, with a temporary deterioration in 1989-90, despite a historically high trade deficit. Since 1990, in particular, the improvement has stemmed both from a slightly improved trade deficit and from stronger invisibles receipts and transfers. EU transfers are also an important factor behind the improvement in the current account in the late 1980s, when they increased substantially. But even if they are excluded, an improvement since 1990 would still be evident, although the size of the current account deficit would be much larger. The share of net EU transfers in the current account deficit has increased since 1987, and reached 4.5 percent of GDP in 1994.

CHART 29
CHART 29

GREECE External Current Account

(In percent of GDP) 1/

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Bank of Greece; Ministry of National Economy; IMF, International Financial Statistics; and OECD, analytical data base.1/ In percent of new GDP; OECD estimates of new GDP were used prior to 1988.
CHART 30
CHART 30

GREECE Composition of the Current Account

(In percent of GDP) 1/

Citation: IMF Staff Country Reports 1995, 095; 10.5089/9781451816051.002.A001

Source: Bank of Greece; Ministry of National Economy; IMF, International Financial Statistics; and OECD, analytical data base.1/ In percent of new GDP; OECD estimates of new GDP were used prior to 1988.

b. Export performance

This section analyzes Greece’s export performance in some more detail, by looking into the commodity composition and destination of Greece’s exports on the one hand, and by comparing Greece as an exporter with a control group of three other EU countries on the other. This analysis adds important insights to the assessment of Greece’s competitiveness that emerges from the discussion in the previous sections.

As shown in the Table below, Greece’s exports during 1980-93 have remained concentrated in traditional products, i.e., primary commodities and labor-intensive goods. These goods accounted for about two-thirds of the total export growth of Greece to the rest of the world. Perhaps more significantly, Greece’s exports to the EU--Greece’s most important trading partner--have also concentrated in these traditional products. Furthermore, Greece’s reliance on this type of exports has not declined over time: in the second half of the 1980s, the contribution of primary and labor intensive products to total export growth was virtually the same as in the first half.

Contribution to Export Growth Rates 1/

article image
Source: TARS data base, IMF.

Primary products include food, beverages, tobacco, mining, oil, and fuel (SITC: 0-4). Natural resource-intensive manufactures include leather, cork products, and paper (SITC: 61-65). Capital- or medium-technology-intensive manufactures include metal, and nonferrous metals (SITC: 66-69). Labor-intensive manufactures include lightning equipment, furniture, travel goods, clothing, and footwear (SITC: 81-85, 89). High-technology-intensive manufactures include camera, film, machinery, chemicals and plastics (SITC: 5, 7, 87, 88). Others covers products belonging to SITC: 9.

Comparing Greece’s export performance against a control group of similar countries reinforces these results. For this comparison, the constant-market-share (CMS) method is used (Learner and Stern 1970; Richardson 1971). The CMS approach decomposes the growth of exports of a particular country into two parts: the growth of exports of the control group; and changes in the country’s relative export performance vis-à-vis the control group. The first part can, in turn, be further decomposed into three elements: the control group growth effect, the commodity composition effect, and the market distribution effect. The second part is commonly interpreted as reflecting changes in the country’s competitive position against the control group.

The main caveat of this approach is the arbitrary selection of the control group members. For purposes of analyzing Greece’s export performance in this Chapter, three other low-income EU countries were chosen as the other members of the control group: Spain, Portugal, and Ireland. The group of these four countries is hereafter denoted as EU-4.

Formally, the decomposition of export growth into the four factors described above can be expressed as:

( X t + 1 − X t ) / X t = ( 1 ) r + ( 2 ) [ Σ i ( r i − r ) ⋅ X i , t ] / X t + ( 3 ) [ Σ i Σ j ( r i j − r i ) ⋅ X i j , t ] / X t + ( 4 ) [ Σ i Σ j ( X i j , t + 1 − X i j , t − r i j ⋅ X i j , t ) ] / X t

where:

Xt: Greece’s total exports in period t.

Xi,t: Greece’s exports of commodity i in period t.

Xij,t: Greece’s exports of commodity i to country j in period t.

r: percentage increase in total EU-4 exports from t to t+1.

ri: percentage increase in the EU-4 exports of commodity i from t to t+1.

rij: percentage increase in the EU-4 exports of commodity i to country j from t to t+1.

The commodity composition effect for commodity i consists of two parts: (ri - r) and Xi,t. If Greek exports of commodity i increased faster than total exports of the EU-4 region, the term (ri - r) would be positive. The relative importance of this term for the commodity composition effect would be large if the other term, Xi,t, is large relative to total exports. Accordingly, the sum, [Σi(ri - r)⋅Xi,t], would be positive if Greece’s largest exports show relatively high growth. The market distribution effect [ΣiΣj(rij - ri)⋅Xij,t] would be positive if Greece concentrated its exports in markets that experienced relatively rapid growth. Finally, the residual effect reflects the country’s relative export performance vis-à-vis the control group. It should be noted that, although it is usually referred to as the “competitiveness effect”, this residual item is also affected by factors other than competitiveness, such as the development of new export products.

The table below shows the result of the CMS analysis, comparing Greece’s exports in selected years (1982, 1984, 1989, and 1993) against 1980 as a base year. Ten SITC commodity groups and 8 market groups (j=1…8) were specified.

Constant-Market-Share Analysis

article image

The export growth rates for 1980-81, 1980-82, 1980-84 were negative. Since the negative export growth rates were set at 100 during this period, the positive (negative) sign of each effect indicates a negative (positive) contribution to the negative export growth.

The table reveals that the control group growth effect was the driving force for Greece’s export growth throughout the period, except in the early 1980s, when the world recession brought about negative export growth. The negative commodity composition effect suggests that Greece’s exports were predominantly goods that experienced low market growth. The growing share of this effect is explained by Greece’s further specialization in its traditional products after its accession to the EU. The negative market distribution effect indicates that Greece concentrated its exports in regions that experienced relatively slow growth. Finally, the large negative residual effect shows that Greece’s performance as an exporter lagged behind that of the other countries in the EU-4 group.

4. Conclusion

The picture that arises from the analysis of both indirect and direct indicators of Greece’s competitiveness is mixed. Price competitiveness has deteriorated since the mid-1980s, but at least some of this is the effect of a rise in the relative price of nontradable goods. This development, in turn, common in many other European countries, is not necessarily an indication of deteriorating competitiveness of Greece’s tradable goods sector. Indeed, cost competitiveness and, in particular, cost competitiveness in manufacturing--the tradable sector par excellence--appears to be stronger now than it has been on average during the last fifteen years or so, and earlier export market share gains have been maintained. Moreover, Greece’s external position has strengthened: the deficit of the current account of the balance of payments reached an all-time low in 1994 and is projected to deteriorate only marginally in 1995, despite the recovery.

There are also, however, worrying signs arising from the detailed analysis of Greece’s export performance. Greece’s exports continue to be concentrated in traditional goods, in which competition has pushed profitability down and is likely to continue doing so in coming years. And Greece’s ability to diversify and expand its exports during the 1980s and early 1990s compares unfavorably with that of other similar countries, like Spain or Portugal.

APPENDIX: Fund Relations with Greece

(As of June 30, 1995)

I. Membership status

Greece is an original member of the Fund (December 27, 1945). Greece accepted the obligations of Article VIII, Sections 2, 3, and 4 of the Fund Agreement on July 22, 1992.

II. General Resources Account:

article image

III. SDR Department:

article image

IV. Outstanding Purchases and Loans: None

V. Financial Arrangements: None

VI. Projected Obligations to Fund: None

VII. Exchange Rate Arrangements

A managed float of the drachma has been in operation since 1975 with the exception of the period January-August 1983, when the drachma was pegged to the U.S. dollar. Although the drachma is included in the ECU and the Bank of Greece has concluded swaps of 20 percent of its gold and U.S. dollar holdings against the ECU with the European Monetary Institute, Greece is at present not a participant in the exchange rate mechanism of the European Monetary System. The drachma/U.S. dollar rate is fixed daily in the light of market conditions and official quotations for other currencies are determined daily on the basis of the official rates between the U.S. dollar and the currencies concerned.

VIII. Article IV Consultation

Greece returned to a standard 12-month consultation cycle in 1993. The last Article IV consultation discussions were concluded on July 18, 1994 (EBM/94/64).

IX. Technical Assistance:

article image

X. Resident Representative: None

References

  • Aghevli, B., J. Boughton, P. Motitiel, D. Villanueva, and G. Woglom (1990), The Role of National Saving in the World Economy, IMF Occasional Paper No. 67, March, Washington D.C.

    • Search Google Scholar
    • Export Citation

  • Aschauer, D. (1985), Fiscal Policy and Aggregate Demand, American Economic Review, Vol. 75, no. 1 (March), pp. 117–27

  • Auerbach, A.J. (1982), Index of Leading Indicators: “Measurement Without Theory,” Thirty-Five Years Later. Review of Economics and Statistics, 64:589-95, November 1982

    • Search Google Scholar
    • Export Citation

  • Bailey, M. J. (1971), National Income and the Price Level, McGraw-Hill

  • Balassa, Bela (1964), “The Purchasing Power Parity Doctrine: A Reappraisal,” Journal of Political Economy, Vol. 72, pp. 584–96

  • Baxter, M. and M. Crucini (1993), Explaining Saving-Investment Correlations, American Economic Review, Vol. 83, no. 3, (June), pp. 416–36

    • Search Google Scholar
    • Export Citation

  • Bayoumi, T. (1990), Saving-Investment Correlations: Immobile Capital, Government Policy, or Endogenous Behavior?, IMF Staff Papers, Vol. 37, no. 2, pp. 360–85

    • Search Google Scholar
    • Export Citation

  • Beckman, B.A., and Tapscott, T.R. (1987), Composite Indexes of Leading. Coincident. and Lagging Indicators. Survey of Current Business, Bureau of Economic Analysis, U.S. Department of Commerce, 67:24-28, November 1987

    • Search Google Scholar
    • Export Citation

  • Boehm, E.A. and Moore, G.H. (1987), New Economic Indicators for Australia, 1949-84, Australian Economic Review, No. 68:34-56, 4th quarter, 1984

    • Search Google Scholar
    • Export Citation

  • Cerea, G. (1982), Public Expenditure and Decisions on Private Consumption, Public Finance, Vol. 37, no. 3, pp. 34–60

  • Feldman, Robert A. (1994), Measures of External Competitiveness for Germany IMF Working Paper WP/94/113

  • Feldstein, M. and C. Horioka, (1980), Domestic Saving and International Capital Flows, The Economic Journal, Vol. 90 (June), pp. 314–29

    • Search Google Scholar
    • Export Citation

  • Fieleke, N. (1982), National Saving and International Investment, in Saving and Government Policy, Federal Reserve Bank of Boston Conference Series No. 25, Boston, Massachusetts

    • Search Google Scholar
    • Export Citation

  • Giovannini, A. (1983), The Interest Elasticity of Savings in Developing Countries: The Existing Evidence, World Development, Vol. 11, no. 7, pp. 601–7

    • Search Google Scholar
    • Export Citation

  • Gorton, G. (1984), Measuring Economic Activity with the Indexes of Leading, Lagging, and Coincident Indicators, Federal Reserve Bank of Philadelphia, Working Paper No. 84-6

    • Search Google Scholar
    • Export Citation

  • Hamann, A. J. (1993), Private Saving. Public Saving and the Inflation Tax: Another Look at an Old Issue, IMF Working Paper No. 93/37

  • IMF (1995), World Economic Outlook, Washington DC, (May)

  • IMF (1994), Exchange Rates and Economic Fundamentals: A Framework for Analysis, Occasional Paper 115

  • IMF Paper on Policy Analysis and Assessment, PPAA No.93/5

  • IMF (1984), Issues in the Assessment of the Exchange Rates of Industrial Countries, Occasional Paper 29

  • Karras, G. (1994), Government Spending and Private Consumption: Some International Evidence, Journal of Money, Credit, and Banking, Vol. 26. no. 1 (February)

    • Search Google Scholar
    • Export Citation

  • Kenen, Peter B. and Claire Pack (1980), exchange Rates, Domestic Prices, and the Adjustment Process: A Survey of Recent Evidence, Occasional Papers No. 1, Group of Thirty: N.Y

    • Search Google Scholar
    • Export Citation

  • Learner, E.E. and R.M. Stern (1970), Quantitative International Economics, Boston: Allyn and Bacon

  • Lemmen, J. and S. Eijffinger (1993), The Degree of Financial Integration in the European Community, De Economist, Vol. 141, no. 2, pp. 189–213

    • Search Google Scholar
    • Export Citation

  • Lipschitz, Leslie and Donogh McDonald (1991), Real Exchange Rates and Competitiveness: A Classification of Concepts, and Some Measurements for Europe, IMF Working Paper, WP/91/25

    • Search Google Scholar
    • Export Citation

  • Marsh, Ian W. and Stephen P. Tokarick (1994), Competitiveness Indicators: A Theoretical and Empirical Assessment, IMF Working Paper WP/94/29

    • Search Google Scholar
    • Export Citation

  • McCallum, J. (1992), Government Spending and National Saving, in M. Baldassari et al (eds.) Deft, Deficit and Economic Performance, St. Martin’s Press, New York

    • Search Google Scholar
    • Export Citation

  • Mitchell, W.C., and Burns, A.F. (1938), Statistical Indicators of Cyclical Revivals, New York, National Bureau of Economic Research

  • Moore, B. (1993), A Review of CSO Cyclical Indicators, Economic Trends No. 447, July 1993

  • Murphy, R. (1984), Capital Mobility and the Relationship between Saving and Investment Rates in OECD Countries, Journal of International Money and Finance, Vol. 3, pp. 327–42

    • Search Google Scholar
    • Export Citation

  • Richardson, J. David (1971), Constant-Market Shares Analysis of Export Growth, Journal of Economic Economics 1, pp. 227–239

  • Seater, J. (1993), Ricardian Equivalence, Journal of Economic Literature, Vol. 31 (March), pp. 142–90.

  • Stock, J.H. and Watson, M.W. (1993), A Procedure for Predicting Recessions with Leading Indicators: Econometric Issues and Recent Experience, in Stock, J.H. and Watson, M.W., Business cycles, indicators, and forecasting, University of Chicago Press, 1993

    • Search Google Scholar
    • Export Citation

  • Stock, J.H. and Watson, M.W. (1988), A Probability Model of the Coincident Economic Indicators, National Bureau of Economic Research, Working Paper No. 2772, November 1988

    • Search Google Scholar
    • Export Citation

  • System of National Accounts (1993), Washington, D.C.

  • Tanzi, V., M. Blejer, and M. Tejeiro (1988), The Effects of Inflation on the Measurement of Fiscal Deficits, in M. Blejer and K. Chu (eds.) Measurement of Fiscal Impact: Methodological Issues, IMF Occasional Paper No. 59 (June)

    • Search Google Scholar
    • Export Citation

  • Tridimas, G. (1992), A Note on the Effects of Government Expenditures on Private Consumption, Public Finance, Vol. 47, no. 1, pp. 153–61

    • Search Google Scholar
    • Export Citation

List of Recent Fund Studies

1994: SM/94/173. Supplement 1 (7/6/94)

Appendix I: Greek Manufacturing 1980-92 in European Perspective

Appendix I: Money Demand in Greece

Appendix III: Exchange and Trade Issues

Appendix IV: Data Issues

1993: SM/93/112 (5/24/93)

Appendix I: Social Security Reforms

Appendix II: Financial Liberalization

1991: SM/91/91 (5/14/91)

Chapter III: Investment and Potential Growth

Chapter IV: Domestic Structural Issues

Chapter V: International Issues

Chapter VI: Inflation and the Private Savings Ratio: Theory and an Application to the Case of Greece

1990: SM/90/99, Supplement 1 (6/11/90)

Public Pension Expenditures in Greece: Long-Term Developments and Prospects

1/

The recovery may actually be stronger than suggested by these estimates. Staff work on a set of coincident indicators of economic activity (see Chapter III) suggests that output growth may have reached 2 percent by end-1994.

2/

As in other European countries, statistical information on trade has become less reliable as a result of the implementation of the Single Market. Underestimation of both exports and imports is likely to occur for various reasons, including the fact that firms are not required to report small transactions.

1/

Over the period 1990-94 as a whole, half of the 6 percent fall in foreign saving reflects an increase in national saving; but most of this has been the result of higher private saving, with government dissaving hovering unchanged at about 10 percent of GDP.

1/

The rate of unemployment based on the number of people registered with the Labor Force Employment Organization (OAED) also remained broadly unchanged: 7.2 percent in 1994, against 7.1 percent in 1993.

2/

There was no change in the rate of employer’s contributions to social security during 1994.

1/

The 30 percent refers to their level in 1989, when unemployment benefits were frozen. The new level will be Dr 2,462, which represents 53 percent of the minimum wage prevailing in January 1994 and 48 percent of the current minimum wage. These figures are below the minimum set by law. While the law stipulates that benefits should be at least two-thirds of the minimum wage, they could be set below that level if OAED’s payments exceed revenues.

1/

The Dr 80 billion debt is the outstanding balance (at the time DEP was separated from the central government) of an account provided by the government from which DEP used to borrow money to purchase oil and repay it at the time of selling the refined products.

2/

See SM/93/112 for a description of the IRO and its operations.

1/

See Gorton (1984) for a survey.

2/

Gorton (1984) gives six criteria used by the Bureau of Economic Analysis of the U.S. Commerce Department: (i) economic significance; (ii) statistical adequacy; (iii) timing at turning points; (iv) conformity to historical business cycles; (v) smoothness; and (vi) timeliness and availability of data.

2/

Boehm and Moore (1987) mention Canada, France, Italy, Japan, the United Kingdom, Germany, Belgium, South Korea, the Netherlands, Taiwan, Sweden and Switzerland.

4/

In addition, SW consider that GDP is a “rather limited” concept where business cycle fluctuations are concerned.

1/

On these grounds, the U.K. Central Statistical Office switched in 1988 to using a GDP-based coincident index.

1/

Or quarter-on-quarter change, when yt is observed quarterly but observable indicators are available monthly.

2/

The choice of equation (6) or (8) depends on whether the unobservable xt is modelled having a unit root or not. A direct test of the order of integration can be performed on yt, but the low power of the usual tests combined with the relatively small number of available annual observations of yt in the case of Greece leave a lot to be desired. For Greece, the Dickey-Fuller statistic is -5.19 with a sample size of 31, which rejects the null hypothesis of a unit root (assuming a non-zero constant and no time trend).

1/

The hypothesis of low international mobility of long-term capital as an explanation of the high correlation between saving and investment does not rely on the existence of capital controls, which in the case of Greece, have now been abolished. But while financial liberalization measures unambiguously increase the mobility of short-term capital, their effect on long-term capital is less clear. Low mobility of long-term capital may be associated with variables that cannot be affected directly by financial liberalization measures, such as specific marketing considerations, tax rules and political risks to which foreign investment may be highly sensitive.

1/

Baxter and Crucini (1993) have argued that a small degree of large-country bias may also affect small countries, to the extent that movements in world interest rates are affected by developments in small countries.

2/

Covers the 12 pre-1995 members.

1/

The data and estimations in this paper are based on the old system of national accounts for the period 1964-1993. The data were taken from various issues of the OECD’s National Accounts series and updated with data provided directly by the authorities. Since the preparation of this study, minor revisions to the data were made available, which are not reflected in the. calculations herein.

2/

Strictly speaking, (T-G) does not represent the government balance (or government saving), because T includes only tax revenues and G includes only consumption expenditures.

1/

For a detailed explanation of the problems created by the inflation component of interest payments see Tanzi et al (1988). For a discussion of the distortions arising from the inflation tax on base see Hamann (1993).

1/

This definition leaves out government paper held by other nongovernment entities for which no information was available at the time of preparation of this study.

1/

The official central government accounts do not include capitalized and accrued interest; according to the SNA methodology, however, these should be included in expenditures (see System of National Accounts, 1993; for a discussion of capitalized interest in Greece see SM/93/112). All figures for interest payments, total expenditures, and overall balances reported herein include capitalized and accrued interest, unless otherwise indicated.

1/

This is a “corrected” figure. The original 1994 budget target was Dr 2,423 billion (10.5 percent of GDP). But the budget presentation in Greece does not include capitalized and accrued interest (equivalent to 1.1 percent of GDP in 1994). In addition, the original 1994 budget excluded payments of interest of Dr 462 billion (2 percent of GDP) due on December 31, 1994, because that day was a holiday. Under proper accrual accounting rules, this amount should have been included in 1994 expenditures. The government has acknowledged this, and in its 1995 budget document has included this amount in the 1994 outcome.

1/

The Ordinary Budget covers all revenues (except EU transfers for investment) and current expenditures.

2/

Account for Guarantees of Agricultural Products, the main vehicle for EU agricultural subsidies to Greece.

3/

Data on central government employment are customarily provided for October 31 of each year.

1/

For a discussion of the different definitions of government in Greek public finance statistics see SM/93/112.

1/

About half of this had been settled in one way or another by the government by end-1994, and the rest was still to be settled.

1/

The official budget target figure (excluding capitalized and accrued interest) is Dr 2,496 billion (9.8 percent of GDP).

2/

The new tax is expected to provide additional revenue of Dr 220 billion, compared with a total personal and corporate income tax revenue of Dr 1,035 billion in 1994.

1/

These issues have been widely discussed in the literature, so only a brief summary of the main points is made here. See, in particular, Tanzi et al. (1988), on which this Chapter is based.

1/

Formally, the rate of depreciation (δ) could be substituted for π in equations (8) and (9). Under PPP (δ-π, assuming zero world inflation) and assumptions (1)-(3), equations (10) and (11) would still be obtained.

2/

In the case of Greece, the stock of domestic debt used excludes drachma-denominated debt indexed to foreign currencies.

1/

The convergence program has been presented and discussed in detail in Suppl. 1 to SM/94/151.

2/

This view is supported by the results of the coincident indicator exercise (Chapter III), which suggest that a stronger-than-officially-estimated recovery had started already in 1994.

1/

M3 includes currency, private sight deposits, time and savings deposits, bank bonds, and repurchase agreements (repos); total drachma financial assets (M4) is M3 plus private sector holdings of government paper with initial maturity of up to one year.

1/

The continuing lack of commercial paper that could be used as collateral has hampered the activation of the rediscount facility.

2/

By comparison, the outstanding stock of overnight interventions in the past was typically less than Dr 150 billion.

3/

In early 1995, M3 growth fell even further, to about 5 percent in the period up to May.

1/

This is a new instrument that minimizes the tax liability on drachma deposits by simultaneously swapping the drachmas for foreign currency--with a lower nominal interest rate and thus a lower tax liability--and entering into a forward agreement to convert the foreign currency deposits back into drachmas at maturity.

2/

M4, however, still does not include holdings of long-term securities, even if these are purchased in the secondary market and have a residual maturity of less than one year, as well as “synthetic swaps.”

1/

For example, between end-1992 and end-1993, interbank rates fell by some 9 percentage points, while the short-term lending rate barely moved. It should be noted that the lending rate reported by the banks is the maximum rate; rates charged to the best customers are 4 percentage points lower.

1/

Although no 12-month bills were issued at the time of the crisis, the government issued one-year bonds in June.

1/

The Agricultural Bank now channels 14.2 percent of its loans toward the non-agricultural sector, up from 1.2 percent in 1991.

2/

Housing loans offered by commercial banks have increased from 5 percent of the total in 1990 to 10 percent of the total in 1994.

1/

The drachma appreciated by 3.8 percent against the U.S. dollar in 1994.

1/

These data probably underestimate exports to Balkan countries, as many transactions with these countries take place in drachmas and are thus not captured in the settlements data.

1/

Prior to this, private external loans with maturity of less than one year were not permitted.

1/

This is not directly comparable to fiscal data.

1/

Measured unit labor costs are highly sensitive to cyclical movements in labor productivity. Normalized ULCs are often used to eliminate these cyclical effects by basing the estimates on trend, rather than actual, productivity (Marsh and Tokarick 1994).

1/

When a country is small so that the terms of trade are exogenously given, exportable and importable goods could be treated collectively as tradable goods. In this case, the analysis could be made on the price difference between tradable and nontradable goods.

1/

These services are: education, housing, health, and public administration.

1/

Data on the current account are available on settlements and on a national accounts basis. For a discussion of the differences between the two data sources, see SM/93/112.

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Greece: Background Paper
Author:
International Monetary Fund
Volume/Issue:
Volume 1995: Issue 095
Publisher:
International Monetary Fund
ISBN:
9781451816051
ISSN:
1934-7685
Pages:
171
DOI:
https://doi.org/10.5089/9781451816051.002
CHART 1

GREECE Comparisons of Economic Performance
CHART 2

GREECE Contributions to Growth 1/

(In percent)
CHART 3

GREECE: Wholesale and Consumer Prices

(Percent changes)
CHART 4

GREECE Real Wages and Unit Labor Costs in Manufacturing
CHART 5

GREECE Coincident Indicators I

(Logarithmic index, end-1985 equals zero)
CHART 6

GREECE Coincident Indicators II

(Logarithmic index, end-1985 equals zero)
CHART 7

GREECE XL1: Actual vs Estimated Output

(Year-on-year logarithmic change)
CHART 8

GREECE XD1: Actual vs Estimated Output

(Year-on-year logarithmic change)
CHART 9

GREECE XL1: Actual vs Estimated Output (level)

(In real 1985 drachma billion)
CHART 10

GREECE European Union: Investment and Saving

(In percent of GDP)
CHART 11

GREECE Investment, Private Saving and Government Saving

(In percent of GDP)
CHART 12

GREECE Fiscal Developments

(In percent of GDP)
CHART 13

GREECE Currency Composition of Central Government Debt
CHART 14

GREECE Maturity Structure of Central Government Dept
CHART 15

GREECE General Government Deficits

(In percent of GDP)
CHART 16

GREECE 1994 Inflation Adjusted Deficits: International Comparison

(In percent of GDP)
CHART 17

GREECE Money Growth, Velocity and Inflation
CHART 18

GREECE: Factors Affecting the Growth of Broad Money (M3)

(Flows in billions of drachmas)
CHART 19

GREECE Interest Rates and Differentials

(In percent per annum)
CHART 20

GREECE Exchange Rates and Competitiveness

(1985=100)
CHART 21

GREECE Balance of Payments and Reserves

(Banking settlement basis)
CHART 22

GREECE Private Capital Flows and Overall Balance

(In US$ millions)
CHART 23

GREECE Price-based Real Effective Exchange Rates 1/
CHART 24

GREECE Internal Real Exchange Rates

(Relative price of nontradable to tradable goods)

(1980=100)
CHART 25

GREECE Cost-based Real Effective. Exchange Rates 1/
CHART 26

GREECE Profitability
CHART 27

GREECE Export Performance

(1970=100)
CHART 28

GREECE Comparison of Export Performance

(1980=100)
CHART 29

GREECE External Current Account

(In percent of GDP) 1/
CHART 30

GREECE Composition of the Current Account

(In percent of GDP) 1/