This Selected Issues paper explores the determinants of Greece’s potential output, an exercise that could shed light on its current cyclical position, as well as on its medium-term growth prospects. To that end, a production function is estimated, with capital and labor as factors of production and time-varying total factor productivity. The estimation results suggest that the output gap has been closing in recent years, with output projected to be slightly above potential in 1998. The paper also explores the impact of changes in the drachma’s exchange rate on inflation.

Abstract

This Selected Issues paper explores the determinants of Greece’s potential output, an exercise that could shed light on its current cyclical position, as well as on its medium-term growth prospects. To that end, a production function is estimated, with capital and labor as factors of production and time-varying total factor productivity. The estimation results suggest that the output gap has been closing in recent years, with output projected to be slightly above potential in 1998. The paper also explores the impact of changes in the drachma’s exchange rate on inflation.

I. Potential Output and the Output Gap in Greece: a Third Face for Janus?1

A. Introduction

4. In a recent paper, Alogoskoufis (1995) argued that, like the Roman god Janus, the post-war Greek economy had displayed two faces. Prior to 1974, economic growth was second only to Japan among OECD members, with inflation rates that were at or below those of its partners. Subsequently, in contrast, the growth rate collapsed, and inflation shot up to 15–20 percent and remained, until recently, stubbornly high. Since 1994, growth has again picked up and inflation moderated, although it remains unclear whether this portends a “third face” for the Greek economy.

5. This chapter presents estimates of potential output and the output gap for Greece, as a means to gauge the appropriateness of the stance of macroeconomic policies, and to assess the prospects for real convergence in living standards with those of EU partners. It is of course necessary to bear in mind throughout that all estimates of potential output are subject to significant margins of uncertainty, and need to be supplemented by judgmental considerations. In the specific case of Greece, in addition, data problems (e.g., affecting the capital stock and the labor force survey) compound the traditional difficulties, as do significant regime shifts in the last few decades (including changes in the political system and related policy approaches, membership in the European Community, etc.).

6. The chapter is organized as follows. Section B discusses the method used for estimating potential output and the output gap, namely, the production function approach, and contrasts it with a simple univariate smoothing procedure. Section C discusses measures of potential output and the output gap for Greece, and examines the contributions made by the various productive factors. Section D focuses on the history of and prospects for the growth in total factor productivity, and offers some tentative econometric evidence on factors that may have influenced its performance and that may determine its future evolution. Section E summarizes the main findings.

B. The Production Function Approach

7. There are a variety of approaches to estimating potential output and the output gap, each with its own strengths and weaknesses.2 One approach that is often used is to smooth the real output series using a Hodrick-Prescott (HP) filter. This procedure minimizes a weighted combination of the gap between actual and trend output and the rate of change of trend output over the sample period. It has the advantage of simplicity, but a number of drawbacks. First, the weight given to minimizing the gap relative to minimizing the variation in trend output is arbitrarily chosen. Second, the technique results in edge-sample biases, although the use of medium-term growth projections can help to reduce the end-sample bias.

8. The approach taken in this chapter is to employ a production function.3 In its simplest form, output is accounted for by a combination (in a Cobb-Douglas functional form) of labor and capital, as well as the state of technology at each point in time, as follows:

Ln(Y)=αLn(L)+(1α)Ln(K)+tfp(1)

where Y, L and K are real gross domestic product, labor, and the nongovernment capital stock, respectively, α is labor’s share in value added and tfp is total factor productivity (in log form).4 Total factor productivity is derived residually by subtracting the weighted contributions of labor and capital.5 Potential output is computed as:

Ln(Y*)=αLn(L*)+(1α)Ln(K)+tfp*(2)

where tfp* is trend total factor productivity obtained by smoothing the residuals of equation (1) using the HP filter.6 L* is the trend labor input calculated as:

L*=LF*(1UR*)(3)

where LF* and UR* are the trend labor force and unemployment rate, respectively, obtained by smoothing the underlying series using the HP filter.

9. One advantage of the production function approach compared to univariate detrending is that it can “explain” the changes in potential output in terms of changes in its underlying components. It is also preferable to univariate smoothing when examining prospects for potential growth, as it can incorporate anticipated changes in the underlying components (e.g., changes in the labor force participation rate). But it has a number of drawbacks as well. First, the estimates may be sensitive to edge-sample biases of the components, although use of medium-term forecasts may again mitigate this problem. Second, the estimates may also be subject to measurement errors in factor inputs, especially the capital stock.

C. Measures of Potential Output and the Output Gap

10. Estimates of potential output, its growth rate, and the output gap for Greece are presented in Figure 1.7 The average rate of potential growth has fallen from about 7 percent in the 1960–73 period, to less than 2 percent since the early 1980s.8 Growth in activity has picked up in recent years, exceeding 2 percent since 1994, and has outpaced the estimated rate of potential growth, although the latter has risen as well. As a result, the output gap is estimated to have largely closed in 1997, with activity set to be slightly (about ¼ percentage point) above potential in 1998. The EU Commission projects negligible slack in 1998, while the OECD estimates an output gap of slightly over 1 percent as persisting in the current year (Table 1).9 The staff and the EU Commission both project output to exceed potential in 1999 (by about ¾ percent), while the OECD projects a gap of ½ percent of potential GDP.

Figure 1.
Figure 1.

Greece: Actual and Potential Output, and the Output Gap 1/

Citation: IMF Staff Country Reports 1998, 100; 10.5089/9781451816082.002.A001

Sources: OECD Analytical Database; MNE; and staff calculations.1/ Shaded areas show staff projections.
Table 1.

Greece: Actual and Potential Growth, and the Output Gap

article image
Sources: OECD Economic Outlook, June, 1998; EU Commission, Spring 1998 forecasts; and Fund staff calculations.

Percent of potential output.

12. What has accounted for the sharp decline in the rate of potential output growth in Greece? The dominant factor, accounting for 4 percentage points of the 5¼ percentage point drop in the potential growth rate between 1960–73 and 1982–97, was a decline in the rate of growth of total factor productivity (Table 2; and Figure 2), which is examined in the next section.10 As regards the reduced contribution to growth from productive factors (accounting for the remaining 1 percentage point), this in turn was the result of an increase in labor’s weighted contribution (from negative initial values) that was more than offset by a reduction in the weighted contribution of capital. Labor’s negative contribution during 1960–73 reflects declining labor force participation rates due to the movement away from agriculture (where women are all assumed to participate in the labor force) and increasing incomes and broadened social insurance programs that allowed for earlier retirement ages and longer education periods prior to entering the labor force (Figure 3). While the male participation rate continued its secular decline in later years, female participation rates rebounded, especially after 1980, in part reflecting changing public attitudes.11

Table 2.

Sources of Growth, International Comparisons

article image
Sources: Austria, forthcoming IMF Country Report; Iceland, IMF Country Report No. 97/15; Switzerland, IMF Country Report No. 97/18; United Kingdom, IMF Country Report No. 96/130; United States, IMF Country Report No. 96/93; Chile, WP/97/104; Korea and Taiwan Province of China, A. Young (1995); and Fund staff calculations.

Business sector.

Non-fish business sector.

Figure 2.
Figure 2.

Greece: Potential Output and Component Growth 1/

Citation: IMF Staff Country Reports 1998, 100; 10.5089/9781451816082.002.A001

Sources: Data from OECD Analytical Database and MNE and staff projections and calculations.1/ Shaded areas show staff projections.
Figure 3.
Figure 3.

Greece: Labor Force Developments

Citation: IMF Staff Country Reports 1998, 100; 10.5089/9781451816082.002.A001

Source: OECD; Tsaliki (1991); and MNE.

13. With regard to the contribution of capital, the decline in the pace of investment in Greece after the mid-1970s (Figure 4) is well documented. Tsaliki (1991) argued that the drop off in investment was the result of negative accelerator effects from slower growth rates for overall activity, a rising relative price of investment goods compared to total output, and a sharp decline in the rate of return to investment after 1973. Alogoskoufis (1995) found a similar pattern in the shadow price of capital, as proxied by the stock market index deflated by the investment goods deflator. As discussed more fully in the next section, this was seen to reflect increasing distortions due to changes in labor market behavior, rising unit labor costs, a shift in the tax regime against capital, concerns about the general protection of property rights, and a host of other rigidities. It was not until after financial sector liberalization, initial reforms of the labor market (notably in the wage indexation and bargaining system), and other supporting structural reforms were introduced in the late 1980s-early 1990s, and with a more continuous pursuit of stability-oriented macroeconomic policies, that private sector investment began to revive.

Figure 4.
Figure 4.

Greece: Non-Government Investment and Capital Ratios 1/

Citation: IMF Staff Country Reports 1998, 100; 10.5089/9781451816082.002.A001

Sources: Data from OECD Analytical Database and MNE and staff projections and calculations.1/ Shaded areas show staff projections.

14. One may question how sensitive the results are to variations in the initial capital stock, as these data were generated based upon the initial (1960) levels of investment. A reduction in the nongovernment investment/potential GDP ratio by one standard deviation (equivalent to 6.8 percentage points of potential GDP) reduces the initial government capital stock/potential GDP ratio by 31.4 percentage points of potential GDP. However, given the subsequent historical levels of investment, it initially increases the growth rate in nongovernment capital in subsequent years, thereby increasing its weighted contribution to growth (Table 3). These effects diminish as the difference in initial capital stocks are subject to depreciation.12 While the contribution of tfp to potential growth was correspondingly slightly reduced in 1960–73, its subsequent collapse continued to account for the bulk of the change in potential growth.

Table 3.

Greece: Sources of Growth

article image
Sources: Data from OECD Analytical Database and MNE; and Fund staff projections and calculations.

D. The Role of Total Factor Productivity

15. Productivity was almost stagnant over 1982–97, growing by far less than that in a sample of other economies for which recent comparable studies are available (Table 2). Identifying concrete reasons for the sharp drop in tfp growth is not straightforward. By construction, the term is a residual, in some sense a measure of our difficulty in explaining the growth process. Before examining factors that may have directly influenced the growth of tfp, it could mechanically be indirectly affected through a more careful estimation of the contributions made by labor and capital. This could include, for example, taking into account the effects of variations in education and hours worked on labor’s effective contribution, and in capacity utilization for capital’s contribution. Generally speaking, improvements in education levels tend to increase the effective labor input and to reduce the size and influence of tfp. This effect appears, however, to have been relatively small in Greece, and a further analysis was therefore not undertaken in this study.13

Factors affecting tfp growth

16. The fact that tfp is determined residually is not to say that we have do not have any ideas about factors aside from labor and capital that may influence the growth process, simply that they may be hard to quantify.14 In this section we explore three potential factors, namely government capital, inflation, and the extent of state intervention in the economy.

17. A great deal of research has been undertaken to ascertain the role that public investment plays in facilitating economic growth. Aschauer (1989a) found a strong, positive relationship between public capital and productivity for the United States, and, in Aschauer (1989b), he extended these results to all of the G-7 economies. Baxter and King (1993) calibrated a small theoretical model to U.S. data and found a strong link between public investment and growth. In a cross-section data set for over 200 economies in the 1970s and 1980s, Easterly and Rebelo (1993) found that public transport and communications investments were significantly positively correlated with growth in per capita income. Nevertheless, the debate concerning the role of public infrastructure in productivity is far from settled. In a survey of this debate, Gramlich (1994) cites econometric difficulties regarding nonstationarity of the underlying series, possible simultaneity biases, and the economic implausibility of the large size of the estimated coefficients for public infrastructure.

18. The pattern of government capital accumulation in Greece reflects in some respects the pattern of tfp growth. In comparison to the variation in the private sector investment/potential GDP ratio, government investment patterns, when scaled for their mean values, were even more volatile (Figure 5).15 In contrast to a steady increase in the nongovernment capital/output ratio in Greece, the government investment ratio, combined with the high rate of growth in output, was insufficient to maintain the governmental capital/output ratio.16 Budgetary pressures led to a reduction in the investment ratio in the latter half of the 1970s, and resulted in a further decline in the capital/output ratio. Only with the upturn in investment in the mid-1980s, combined with a collapse in the growth of activity, did the capital/output ratio rebound. However, Alogoskoufis (1995) has argued that these investments focused more on improving social infrastructure (e.g., old-age housing, local hospitals, town squares and pavement projects), and thus were less effective in promoting measured output.

Figure 5.
Figure 5.

Greece: Government Investment and Capital Ratios 1/

Citation: IMF Staff Country Reports 1998, 100; 10.5089/9781451816082.002.A001

Sources; Data from OECD Analytical Database and MNE and staff projections and calculations.1/ Shaded areas show staff projections.

19. Beginning in the latter half of the 1980s, the then EC’s structural funds programs took on increasing importance in the context of the European Single Act, with the intention of channeling substantial unilateral transfers to relatively underdeveloped regions of the Community to facilitate convergence in living standards.17 While the initial rate of absorption of potential funds by Greece was poor, it has increased in recent years, with improved project selection, monitoring, and implementation by both the Greek authorities and the EU. A major part of these funds are used for large-scale infrastructure projects, and it is anticipated that public investment will remain sustained in the coming years.18

20. A second factor that has been cited as influencing growth is high and variable inflation (see among others Sarel, 1995; and Ghosh and Phillips, 1998). The channels through which this occurs are potentially numerous (e.g., lower investment resulting from uncertainty due to the positive correlation between high and variable inflation, and the impact of incomplete indexing on real effective tax rates). High inflation may also render uneconomic investments that were prospectively profitable (e.g., from a loss of competitiveness in export sectors given an exchange rate peg). Inflation in Greece has certainly been quite variable, and until recently quite high (Figure 6). It averaged only 2 percent in the 1960s, but jumped to 14 percent in the 1970s and to 19 percent in the 1980s, the period of low tfp growth. It was still above 15 percent as late as 1992, before falling back to 5½ percent in 1997, a period in which tfp growth has picked up.19

Figure 6.
Figure 6.

Greece: Total Factor Productivity and Potential Influences

Citation: IMF Staff Country Reports 1998, 100; 10.5089/9781451816082.002.A001

Sources: OECD Analytical Database; MNE; and staff calculations.

21. A third possible factor that may have affected the growth of tfp is the degree of direct government involvement in economic decision-making. This, of course, is an area where examples of both beneficial and adverse economic influences abound. Following the end of the military regime in Greece in 1974, the democratically elected government was faced with both a political need to undertake social modernization (including labor legislation, social insurance, education reform, and the provision of health care) that had been previously suppressed, and with the economic need to address macroeconomic imbalances.20 General government current primary expenditure rose from an average of 18¼ percent of GDP during 1960–73 to 24¾ percent of GDP during 1974–81, with a further increase to slightly over 31 percent of GDP during 1982–97. Government employment displayed a similar increase, doubling its share in total employment, to 10 percent, between 1960 and 1990 (Figure 6).21 In addition to an activist fiscal policy, direct involvement in economic decisions also increased as successive governments provided support to or nationalized “problem” enterprises, and credit from state-owned banks was channeled to targeted industries or firms. Direct price and interest rate controls, as well as extensive labor market regulations, including distortionary wage indexation policies throughout the 1980s, led to an ossified economic structure and a serious misallocation of resources.22

22. Steps began to be taken in the late 1980s to deregulate segments of the Greek economy, and to reduce the pervasive role of the state in economic decision-making. Financial sector and product market reforms were gradually introduced. As importantly, there was substantial progress toward macroeconomic stability, with sizable reductions in inflation and in the fiscal deficit (although the reliance on increased revenues to achieve the fiscal improvement may have contributed to distortions elsewhere, notably in the labor market).

23. Bringing together the above considerations, an attempt was made to assess the role that government investment, inflation and the degree of government involvement in the economy may have had on tfp growth in Greece. A time series similar to Koedijk and Kremers’ point estimate was not readily available to measure government involvement, so the ratio of government to total employment was used as a crude proxy.23 Attempts were made to find a co-integrating relationship among the rate of growth of tfp, the growth rate of government capital/trend employment, consumer price inflation and the share of government employment over 1961–97 (as all of the series are integrated of order 1):

pch(tfp)=6.71(5.56)+0.22(2.61)pch(Kg/L*)0.08(4.53)infl0.59(5.52)(Lg/L)(4)R2=0.94CRDW=0.88ADF(1)=3.30ADF(4)=2.42

where Kg and Lg are government capital and labor, respectively (t-statistics are in parentheses).

24. The results are rather fragile and should be viewed as merely suggestive of some of the influences on the growth of tfp, but are not implausible (see bottom right panel of Figure 6).24 The coefficient on the percent change in per-worker government capital is not unusually large, as has often been found in other studies (e.g., on the United States, see Gramlich, 1994). The coefficient on inflation has the correct sign and is significant, but is much larger (by a factor of about 10) than those found in Ghosh and Phillips (1998). Finally, the effect of government employment is also found to be significantly negative, and itself is estimated to account for about one-half of the decline in the growth of tfp. Again, while being wary of putting too much weight on these estimates, it is worth noting that the equation tracks the recent upturn in tfp growth quite well.25

25. The above equation has a number of additional shortcomings. First, the tests for nonstationarity of the residuals cannot be rejected at traditional levels of significance, although the co-integration regression Durbin-Watson statistic and the augmented Dickey-Fuller regression with one lag are both significant at the 10 percent level. Second, and more importantly, it is possible that any given set of variables may contain more than one long-run relationship: there may be more than one co-integrating vector. Therefore, the procedure developed by Johansen and Juselius (1990) was followed to test for the number of co-integrating vectors (see appendix for details). This method, while confirming the significant influence of growth in government investment per trend employee and of the share of government employment on the growth in tfp, did not find a significant role for inflation. The equation should thus be viewed as merely suggestive of some of the factors that may have influenced the growth record in Greece.

Prospects for tfp and potential output

26. What are the prospects for the growth of potential output in Greece? The following table compares the staffs projection for growth of potential output and of its components with those implied by forecasts contained in the authorities’ updated convergence program.

article image

27. Turning first to the factors of production, demographic projections entail little change in the growth rate of the population, or of the share of those of working age within the population, in the coming decade (although the latter is expected to fall relatively rapidly thereafter). It is also likely that labor force participation rates will continue to show a medium-term trend increase (despite the reduction that occurred in 1997, which is thought to have reflected sampling problems in the labor force survey), especially as the female participation rate remains far below its EU average. The outlook for unemployment is more problematic. Despite sustained growth in recent years, the unemployment rate has remained above 10 percent of the labor force. Thus, barring significant labor market reforms, it is unclear whether the unemployment rate should be expected to decline.26 Investment prospects, in contrast, are expected to remain buoyant. Aside from the effects of the devaluation of March 1998, a return to disinflation and nominal convergence in Greece is anticipated to lower the cost of capital and improve cash flows. The stock market has risen by some 70 percent since the devaluation, also a sign of improved prospects, and funds raised through the market, while still small compared to other forms of financing, are at all-time highs. It does not appear implausible that the total weighted factor contributions to potential output could be about 2 percent per year, slightly higher than they have been so far this decade.

28. Prospects for the growth of tfp are inevitably more uncertain. At the time of joining the ERM in March 1998, and as articulated further in the updated convergence program, the Greek authorities announced a concerted plan covering privatization, the restructuring of loss-making state-owned enterprises, and labor market and social security reforms. Structural action in these areas will be central to boosting productivity and sustaining a noninflationary growth of employment and investment, as postulated in the convergence program. The privatization program and restructuring of major loss-making enterprises should have both direct as well as demonstration effects on productivity. Extensive guarantees of job security would however be at odds with achieving the required productivity gains. Furthermore, current plans still leave a large state presence in a number of commercial spheres, either because there are no intentions for privatization at this time (including some large commercial banks and traditional utilities), or because the sell-offs are partial, limited to up to 49 percent of each firm. Greater efforts at deregulation and liberalization in sectors that are not presently slated for privatization could help enhance competition and reduce the costs of some basic inputs to the rest of the economy.

29. Regarding social security, the government is adopting a two-stage approach. A first phase consisting of a package of measures that can be quickly implemented is currently under way (it includes registration and required contributions by heretofore illegal immigrant workers, other measures to curtail contribution evasion, merging of various supplementary pension schemes, restricting pensioners’ right to employment to curtail early retirement, and partly liberalizing the choice of investments available for surplus funds). The second, and more ambitious, phase, for which a timetable is to be announced by end-1998, would harmonize key parameters (e.g., minimum retirement ages, contribution periods, and replacement ratios) across all pension schemes, with a view to ensuring the system’s longer-term viability and equity, as well as introduce complementary fully funded schemes. These reforms should, when adopted, improve labor market mobility (with increasing pension portability), and raise labor’s contribution to growth by limiting early retirement and reducing the tax wedge from social security contributions by making part of the system fully funded.

30. The need for comprehensive reforms to improve the functioning of the labor market in Greece is widely recognized.27 Greece has relatively low participation and employment rates, and high youth and long-term unemployment. Strict rules on working time impede the efficient use of labor, while high severance costs and restrictions on collective dismissals deter firms from new hiring. Minimum wages are not differentiated by age or sector, and the overall wage structure is highly compressed (as evidenced by one of the highest ratios of minimum to average production wages in the OECD). The authorities intend to address some of these issues through legislation, to be adopted this year, that would improve work time flexibility, liberalize certain work rules, and allow part-time work in the public sector. Local employment agreements, that would suspend sectoral wage awards in high unemployment areas, are also to be introduced, as are provisions allowing for private (but nonprofit) employment agencies28 Further labor market reforms (to, for example, allow for differentiated minimum wages, reduce firing restrictions and high severance costs, and allow for-profit employment services) would increase the likelihood of improving the growth of potential output.

E. Summary and Conclusions

31. The paper began by asking whether the postwar Greek economy is entering a third phase, characterized by renewed growth that would allow for a sustained convergence in living standards with its EU partners. It was seen that the disappointing growth performance in the decades following the mid-1970s was largely due to a collapse in the growth of total factor productivity, rather than to a reduction in the rate of factor accumulation. While it is comparatively difficult to pin down the reasons for changes in the growth of tfp, it was suggested that variations in the growth of public investment, and in the extent of the government’s direct involvement in economic decision-making, may have played a role, while the effect of inflation on potential growth appears less certain.

32. Recent developments in these key areas portend positively for growth prospects. The pursuit of macroeconomic stability is at the center of the government’s economic policy. The rate of public investment is rising, with increased emphasis on basic infrastructure, supported by an improved absorption of EU funds. In addition, the government has announced a set of structural reforms that should, when fully implemented, allow for increased productivity and flexibility, with scope for additional efforts in this area having the potential to improve prospects further. In all, the projections contained in the new convergence program imply a sharp acceleration in the growth of tfp and of potential output—to rates last seen in the early 1970s. Such a pickup cannot be ruled out, but its realization clearly hinges on carrying forward with a broadly conceived and boldly implemented program of structural reforms.

APPENDIX Results of the Johansen-Juselius Procedure for tfp Growth in Greece

33. In light of the possible econometric shortcomings discussed in Chapter I, the Johansen-Juselius procedure was followed to test for the number of co-integrating vectors. The following table reports the test statistics and the estimated co-integrating vector from the procedure, where r is the number of co-integrating vectors.

Johansen Maximum Likelihood Tests and Parameter Estimates

(1964–1997, maximum of 3 lags in VAR)

article image

The number of co-integrating vectors is denoted by r.

34. Panel A reports the maximal eigenvalue test of the null hypothesis that there are at most r co-integrating vectors against the alternative r + 1 vectors. Beginning with the null hypothesis that there are three or less co-integrating vectors (r ≤ 3) against the alternative of four co-integrating vectors (r = 4), the test statistic (2.59) is less than the 95 percent critical value (3.80), not allowing one to reject the null hypothesis and indicating that there are at most three co-integrating vectors. Similarly, tests for r ≤ 2 against the alternative r = 3 and for r ≤ 1 against the alternative r = 2, cannot be rejected. However, the null hypothesis that there is no co-integrating vector against the alternative that there is one such vector can be rejected at the standard 95 percent significance level, suggesting that there is a unique co-integrating vector. Panel B reports the trace test of the null hypothesis that there are at most r co-integrating vectors against the alternative that there are r + 1 vectors. Again, one cannot reject the null hypotheses that there are three or fewer, two or fewer, and one or fewer co-integrating vectors against their associated alternate hypotheses, but one can reject the hypothesis (at the 99 percent significance level) that there are is no co-integrating vector (r = 0), suggesting that there is at least one co-integrating vector.

35. Panel C of the above table presents the estimated co-integrating vector, with the coefficients normalized on the rate of growth of tfp. The coefficient estimates for the percent change in government capital per trend employee and for the share of government in total employment are appropriately signed, although their values are somewhat higher than those obtained in equation (4) of the main text. The coefficient on inflation, in contrast, is incorrectly signed, although a likelihood ratio test is unable to reject that the variable is not significant, in contrast to the significant t-statistic in equation (4). Thus, from the two approaches, it would appear that the growth in tfp is clearly co-integrated with growth in government capital per worker and with the size of the state in the total economy as measured by its share of employment, but the influence of inflation appears less clear-cut.

References

  • Alogoskoufis, George, 1995, “The Two Faces of Janus: Institutions, Policy Regimes and Macroeconomic Performance,Economic Policy, No. 20, pp. 147192.

    • Search Google Scholar
    • Export Citation
  • Artus, Jacques, 1997, “Measures of Potential Output in Manufacturing for Eight Industrial Countries, 1955–78,Staff Papers, International Monetary Fund, Vol. 24, pp. 135.

    • Search Google Scholar
    • Export Citation
  • Aschauer, David, 1989, “Is Public Expenditure Productive?,Journal of Monetary Economics, pp. 177200.

  • Aschauer, David, 1989, “Back of the G-7 Pack: Public Investment and Productivity Growth in the Group of Seven,Federal Reserve Bank of Chicago Working Paper WP-89–13.

    • Search Google Scholar
    • Export Citation
  • Barro, Robert, and Xavier Sala-i-Martin, 1995, Economic Growth, (New York: McGraw-Hill).

  • Baxter, Marianne, and Robert King, 1993, “Fiscal Policy in General Equilibrium,American Economic Review, No. 3, Vol. 83 (June), pp. 315334.

    • Search Google Scholar
    • Export Citation
  • Christofides, Charalambos, 1996, “Growth Performance: A Survey of the Literature,” in Greece—Recent Economic Developments and Selected Issues, IMF Staff Country Report No. 96/121 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Coe, David, and Dennis Snower, 1996, “Policy Complementarities: The Case for Fundamental Labor Market Reform,IMF Working Paper 96/93 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Curtis, Glenn, 1995, Greece: A Country Study, 4th edition, Federal Research Division, Library of Congress, Washington.

  • Demekas, Dimitri, and Zenon Kontolemis, 1996, “Unemployment: A Survey of the Issues,” in Greece—Recent Economic Developments and Selected Issues, IMF Staff Country Report No. 96/121 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • De Masi, Paula, 1997, “IMF Estimates of Potential Output: Theory and Practice,IMF Working Paper 97/177 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Easterly, William, and Sergio Rebelo, 1993, “Fiscal Policy and Economic Growth: An Empirical Investigation,NBER Working Paper No. 4499.

    • Search Google Scholar
    • Export Citation
  • Garganas, Nicholas, 1992, The Bank of Greece Econometric Model of the Greek Economy, Bank of Greece, Athens.

  • Gaspar, Vítor, and Alfredo Pereira, 1995, “The Impact of Financial Integration and Unilateral Public Transfers on Investment and Growth in EC Capital-Importing Countries,Journal of Development Economics, Vol. 48, pp. 4366.

    • Search Google Scholar
    • Export Citation
  • Ghosh, Atish, and Steven Phillips, 1998, “Inflation, Disinflation, and Growth,IMF Working Paper 98/68 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Giorno, Claude, Peter Richardson, Deborah Roseveare, and Paul van den Noord, 1995, “Estimating Potential Output, Output Gaps and Structural Budget Balance,Economics Department Working Paper No. 152 (Paris: OECD).

    • Search Google Scholar
    • Export Citation
  • Government of Greece, 1998, National Action Plan for Employment: Greece, 1998, Ministry of Labor and Social Security, Athens.

  • Gramlich, Edward, 1994, “Infrastructure Investment: A Review Essay,Journal of Economic Literature, Vol. XXXII, No. 3.

  • Hall, S. G., and N. G. Zonsilos, 1997, “The Output Gap and Inflation in Greece,” mimeo, Athens.

  • International Monetary Fund, 1996, “The Cyclical Position of the U.S. Economy and Its Relationship with Inflation,” in United States—Selected Issues, IMF Staff Country Report No. 96/93 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Jaeger, Albert, 1997, “Estimates of Potential Output Growth and the Cyclical Output Gap,” in Switzerland—Selected Issues and Statistical Appendix, IMF Staff Country Report No. 97/18 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Johanson, Søren, and Katarina Juselius, 1990, “Maximum Likelihood Estimation and Inference on Co-integration—With Applications to the Demand for Money,Oxford Bulletin of Economics and Statistics, Vol. 52, pp. 169210.

    • Search Google Scholar
    • Export Citation
  • Jouganatos, George, 1992, The Development of the Greek Economy, 1950–91: An Historical, Empirical and Econometric Analysis (Westport, Connecticut: Greenwood Press).

    • Search Google Scholar
    • Export Citation
  • Karadeloglou, Pavlos, Christos Papazoglou, and George Zombanakis, 1998, “Is the Exchange Rate and Effective Anti-Inflationary Policy Instrument?,” Bank of Greece Economic Research Department, Athens.

    • Search Google Scholar
    • Export Citation
  • Koedijk, Kees, and Joroen Kremers, 1996, “Deregulation: A Political Economy Analysis,Economic Policy, pp. 443467.

  • Kumar, Monmohan, 1996, “Potential Growth and the Output Gap,” in United Kingdom—Recent Economic Developments, IMF Staff Country Report No. 96/130 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Magnier, Antoine, 1997, “Potential Output and the Output Gap,” in Iceland—Recent Economic Developments, IMF Staff Country Report No. 97/15 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Magnier, Antoine, 1998, “Potential Output and the Output Gap in Austria: A European Perspective,” in Austria—Selected Issues and Statistical Appendix, IMF Staff Country Report (to be published).

    • Search Google Scholar
    • Export Citation
  • OECD, 1996, OECD Economic Surveys: Greece 1996, Paris.

  • OECD, 1998, “The OECD Jobs Strategy: Progress Report on Implementation of Country-Specific Recommendations (Note by the Secretariat),Paris.

    • Search Google Scholar
    • Export Citation
  • Roldos, Jorge, 1997, “Potential Output Growth in Emerging Market Countries: The Case of Chile,IMF Working Paper 97/104 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Sarel, Michael, 1995, “Non Linear Effects of Inflation and Economic Growth,IMF Working Paper 95/56 (Washington: Interntional Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Tsaliki, Persefoni, 1991, The Greek Economy: Sources of Growth in the Postwar Era (New York: Praeger).

  • Vijverberg, Wim, Chu-Ping Vijverberg, and Janet Gamble, 1997, “Public Capital and Private Productivity,The Review of Economics and Statistics, Vol. LXXIX, Number 2, pp. 267278.

    • Search Google Scholar
    • Export Citation
  • Young, Alwyn, 1995, “The Tyranny of Numbers: Confronting the Statistical Realities of the East Asian Growth Experience,Quarterly Journal of Economics, Vol. 110 (3), pp. 64180.

    • Search Google Scholar
    • Export Citation
1

Prepared by Mark Lutz.

2

De Masi (1997) provides an overview of various approaches to estimating potential output that have been used in the Fund. Magnier (1998) includes a discussion of the relative merits of alternate approaches.

3

Artus (1977) was the first IMF research study to adopt the production function methodology to estimate potential output.

4

Data for real gross domestic product and the labor force are taken from the OECD Analytical Database, supplemented in later years by Ministry of National Economy estimates and staff projections. Official capital stock data are not available, and were therefore created using a perpetual inventory method, with the initial stocks based upon the level of investment in 1960, the first year of the sample, on depreciation rates provided by Garganas (1992), and on the average growth of real output during the 1950s (from data included in Tsaliki, 1991). Sensitivity of the sources of growth and of the econometric results to variations in the estimated capital stock are discussed in sections C and D, respectively, below.

5

In the Greek national accounts, the share of dependent labor compensation in output averaged only about one-third during the 1990s, reflecting the large role played by sole proprietorships and the self-employed. The latter sector’s incomes are not separated into returns to labor and capital, but are grouped under “property and entrepreneurial income.” We have assumed that the overall share of income attributed to labor is 65 percent, the lower end of the 65–75 percent range cited by Giorno and others (1995) as typical for OECD countries.

6

The smoothing parameter was set at 25, the same level used by the OECD (see Giorno and others, 1995).

7

Hodrick-Prescott filters were fit over 1960–99, with staff projections used for the last two years to mitigate the end-sample bias.

8

The potential growth rate estimated by the production function and univariate approaches are found to be quite similar. This is because, as discussed more fully in the next section, the bulk of the reduction in potential output growth is due to slower growth in tfp which is simply smoothed by a HP filter in constructing potential output.

9

There are very few national studies of the output gap for Greece. Tsaliki (1991) examined the determinants of potential output during 1950–85, but did not address the output gap. Hall and Zonzilos’ (1997) results covered 1953–95. They used a split-sample technique, with a 1.7 percent growth rate in the final period, 1982–95. This growth rate would imply that the economy was operating more than 3 percentage points above potential in 1997, although it is plausible that their technique would find a new regime for the later years.

10

The dates chosen to break the sample period into subsamples follow those in Zonzilos and Hall (1997).

11

Labor’s contribution to potential growth is also sensitive to the trend unemployment rate. The latter was estimated at 10.1 percent of the trend labor force in 1997, close to the actual unemployment rate of 10.3 percent. The estimated trend rate was also close to the 9.8 percent rate of “structural”, or the nonaccelerating wage rate of, unemployment estimated by the OECD (1998).

12

The difference in the capital stocks was only 10 percent by 1971.

13

In Greece the percentage of workers with no more than six years of education decreased from 42 percent of the workforce in 1961 to 27¾ percent in 1973 and to only 7½ percent in 1991 (the latest data available), while those with at least a secondary school education increased from almost 11 percent in 1961 to 19½ percent in 1973, and to 46½ percent in 1991. Were this by itself taken into account, labor’s effective contribution to growth would be adjusted upward, with a corresponding lower contribution from total factor productivity. Tsaliki (1991) attempted to correct for this and other effects (such as age, gender and hours worked) in her estimates of potential growth over 1950–85, but found, however, that the improvement in education levels was largely offset by reductions in the return to education (which proxied for changes in relative productivity, though this most likely also reflected the indexation system in place in the 1980s that led to a sharp compression of wages). Thus, while these refinements changed slightly the relative contributions of labor, capital and tfp, Tsaliki found that, as in this study, the growth of tfp after 1973 was much lower (by almost 4 percentage points) than it was in the earlier period, and accounted for the bulk of the decline in the growth rate of potential output.

14

See, for example, the suggestions considered in the symposium on the slowdown in productivity growth contained in the fall 1998 issue of The Journal of Economic Perspectives. See Christofides (1996) for a review of these issues in the Greek context.

15

The coefficient of variation for the private sector investment/potential GDP ratio was 0.17, while the corresponding figure for the government investment ratio was 0.30.

16

The change in the capital output ratio is approximately determined by the investment/GDP ratio minus the product of the capital/output ratio, lagged one period, and the sum of the rate of depreciation and the growth rate of the economy. Thus, a higher rate of output growth would, ceteris paribus, reduce the capital/output ratio. However, note that, by comparison, the variation in the government capital/output ratio was quite minor when compared to the nongovernmental ratio.

17

See Gaspar and Pereira (1995) for an analysis of the effect of EU transfers on real convergence of living standards in Portugal.

18

Although at a pace that is likely to be less ambitious than that assumed in the updated convergence program.

19

There remains some dispute about the costs and benefits of low and stable inflation, and of the advantages of further reducing inflation from already low levels. Thus, it would appear that the benefits to the economy from reducing inflation further are not as large as those obtained from bringing inflation down from the high levels experienced in Greece in the 1980s.

20

For a discussion of postwar Greek economic history, see Alogoskoufis (1995), Curtis, ed. (1994), and Jouganatos (1992).

21

Demekas and Kontolemis (1996) found that the rising share of public sector employment and the government’s wage decisions during the 1980s contributed significantly to real wage rigidities and rising unemployment.

22

Koedijk and Kremers (1996) found a clear negative relationship between regulation and economic performance, and characterized Greece as having the most extensive degree of regulation of labor and product markets in the late 1980s-early 1990s among EU economies. Ireland, in contrast, was characterized as having the least regulated labor and product markets, and has been growing this decade at a remarkably brisk pace.

23

One could also use government consumption to GDP, a ratio often used in growth convergence equations; see Barro and Sala-i-Martin (1995). The results were found not to be sensitive to the choice of either variable.

24

One sign of the fragility of the results was that the coefficient on the growth in the government capital stock was found to be significant only at the 10 percent level, and only if lagged one period, while the growth in the government capital stock per trend employee was significant at traditional test levels. This lack of robustness to various specifications was also found in a recent study by Vijverberg, Vijverberg, and Gamble (1997).

25

There is again some question as to the sensitivity of the results to the initial level of the capital stock. A reduction in the investment/potential GDP ratio by one standard deviation (equivalent to 1.3 percentage points) reduced the initial government capital stock/potential GDP ratio by almost 15 percentage points of potential GDP. However, given the subsequent historical levels of investment, it increased the percentage change in the government capital/trend employment by about 2.5 percentage points in 1961, with a smooth convergence in the revised and initial growth rates thereafter. A regression using the lower bound capital stock data resulted in little change in the coefficient estimates or in the overall fit of the equation.

26

In fact, the OECD (1998) projects the unemployment rate to increase further to over 10½ percent this year, and to remain as high in 1999, while the staff foresees a modest decline to 10 percent by next year, close to prevailing estimates of the natural rate.

27

Coe and Snower (1996) make the case for fundamental labor market reform in Europe, stressing the role of complementarities in which changes in policies will have a greater effect on unemployment when adopted jointly, rather than in isolation. As regards Greece, the shortcomings of the functioning of the labor market were highlighted in Demekas and Kontolemis (1996) and in the 1996 OECD Economic Survey for Greece (Chapter 3).

28

The authorities’ intentions in other areas relating to the labor market, including education and training, small businesses formation, and taxation are set out in their recent National Action Plan for Employment (1998).

References

  • Bean, C., 1994, “European Unemployment: A Survey,Journal of Economic Literature, Vol. 32, pp. 573619.

  • Christofides, C., 1995, “An Index of Coincident Economic Indicators for Greece,” in Greece—Background Paper, IMF Staff Country Report No. 95/95 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Fabiani, S., A. Locarno, G. Oneto, and P. Sestito, 1998, “Risultati e Problemi di un Quinquennio di Política dei Redditi: Una Prima Valutazione Quantitativa,” Banca d’Italia, Temi di Discussione No. 329.

    • Search Google Scholar
    • Export Citation
  • Ford, R. and T. Krueger, 1995, “Exchange Rate Movements and Inflation Performance: The Case of Italy,IMF Working Paper 95/41, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hall, S. G. and N. G. Zonzilos, 1996, “The Output Gap and Inflation in Greece,” mimeo, Bank of Greece.

  • Karadeloglou, P., C. Papazoglou, and G. Zombanakis, 1997, “Is the Exchange Rate an Effective Anti-Inflationary Policy Instrument?,” mimeo, Bank of Greece.

    • Search Google Scholar
    • Export Citation
  • Layard, R., S. Nickell, and R. Jackman, 1991, Unemployment (Oxford University Press).

  • Litterman, R. and L. Weiss, 1985, “Money, Real Interest Rates, and Output: A Reinterpretation of Postwar U.S. Data,” Econometrica, 53.

    • Search Google Scholar
    • Export Citation
  • Lucas, R. E., Jr., 1976, “Econometric Policy Evaluation: A Critique,” in The Phillips Curve and Labor Markets, ed. by K. Brunner and A. H. Metzler (New York: North-Holland).

    • Search Google Scholar
    • Export Citation
  • Obstfeld, M., 1997, “Europe’s Gamble,Brookings Papers on Economic Activity, 2.

  • OECD, 1997, “Recent Labor Market Developments and Prospects,” Chapter I, Employment Outlook.

  • Oxford Economic Forecasting, 1995, “New Oxford World Model,” Oxford.

  • Sims, C. A., 1980, “Macroeconomics and Reality,” Econometrica, 48.

  • Sims, C. A., 1982, “Policy Analysis with Econometric Models,Brookings Papers on Economic Activity.

  • Sobczak, N., 1998, “Disinflation in Spain: the Recent Experience,” in Spain—Selected Issues, IMF Staff Country Report No. 98/53 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

STATISTICAL APPENDIX

Table 4.

Greece: Aggregate Demand

(At constant prices of the previous year)

article image
Sources: Ministry of National Economy; and Fund staff calculations.
Table 5.

Greece: Aggregate Demand

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

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.

Table 7.

Greece: Saving-Investment Balance

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

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

Current account deficit.

Table 8.

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 9.

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.