This Selected Issues paper on Korea provides background information on economic developments and policies, with particular emphasis on 1995–96. Following two years of rapid expansion, led by buoyant investment and exports, economic growth moderated in late 1995 and the first half of 1996. The moderation was in response to the earlier tightening of monetary conditions and less favorable short-term export prospects. The slowdown was reflected in a sharp deceleration in final domestic demand, whose contribution to growth fell from 9.1 percent in 1995 to 6.6 percent in the first half of 1996.

Abstract

This Selected Issues paper on Korea provides background information on economic developments and policies, with particular emphasis on 1995–96. Following two years of rapid expansion, led by buoyant investment and exports, economic growth moderated in late 1995 and the first half of 1996. The moderation was in response to the earlier tightening of monetary conditions and less favorable short-term export prospects. The slowdown was reflected in a sharp deceleration in final domestic demand, whose contribution to growth fell from 9.1 percent in 1995 to 6.6 percent in the first half of 1996.

V. Potential Output Growth: Past Experience and Prospects 1/

1. Introduction

Estimates of potential output play an important role in the staff’s analysis of macroeconomic developments in Korea. Based on the assumption that potential output growth is a relatively smooth process with short-run fluctuations in growth largely the result of variations in demand conditions, 2/ the ratio of actual to potential output (output gap) can be used as an indicator of resource pressure and thus of prospective inflation developments. In addition, estimates of potential output are used to calculate many macroeconomic indicators, such as the cyclically adjusted fiscal and current account positions, and for medium-term scenarios addressing issues of balance of payments and debt sustainability. The concept of potential output underlying these applications is an economic rather than a technical one. It refers to the level of output consistent with a constant rate of inflation and unemployment equal to its natural rate. For an economy that is close to the level of potential, the growth rate of potential output indicates the rate at which it can expand over the medium-term without risking a rise in inflation.

Beyond its application in the staff’s assessment of macroeconomic developments and policies, the analysis of potential output growth also provides useful insights into the roles of factor input and productivity growth in Korea’s past growth experience, which has been extensively debated in recent years in the wider context of the discussion on the “East Asian Miracle.” 3/ Looking ahead, the key question is whether, or rather under which conditions, the high growth rates experienced in the past can be sustained.

This chapter presents new estimates of potential output for Korea, 4/ revisits the debate on the nature of the rapid growth in the past 25 years, and discusses the outlook for potential output growth during the next decade under alternative assumptions about developments in the factors influencing the growth of labor and capital inputs and total factor productivity. Section 2 describes the framework used to estimate potential output and the main estimation results. Section 3 examines the historical growth contributions of labor, capital, and productivity. Section 4 discusses alternative scenarios for potential growth in the next decade, and Section 5 summarizes the results.

2. Estimation framework and results

A number of different approaches have been used to estimate potential output. 1/ Univariate methods such as simple trend estimates, moving averages, or more sophisticated filtering procedures 2/ generally yield useful estimates of output gaps for the short-run analysis of inflation developments or the calculation of various macroeconomic indicators, and are thus widely used. 3/ These methods provide, however, no insight into the underlying determinants of growth. Moreover, they allow projections of future potential output growth only on the basis of an extrapolation of past trends, which may become less and less relevant as the projection period lengthens. A production function approach, by contrast, permits an assessment of the growth contributions of labor, capital and total factor productivity, as well as projections of future potential growth on the basis of alternative assumptions about the determinants of factor inputs and productivity.

The estimates of potential output discussed in this chapter are based on a simple Cobb-Douglas production function, which relates output Y to capital and labor inputs, K and L, as well as total factor productivity A

Y=A·Kα·Lβ(1)

with α + β = 1 under the assumption of linear homogeneity, and the variables Y, K, and L indicating potential levels. Actual output Ya is then described by

Ya=A·U·Kα·(LU·L)β(2)

where U indicates the intensity of overall factor utilization, which varies over the business cycle, and LU a specific utilization factor for L. This formulation allows for differences in the intensity of utilization of K and L, and permits the incorporation in the estimating equation of specific information on the utilization of labor input, such as data on unemployment rates and cyclical variations in participation rates. If L is defined as the level of labor input when the unemployment rate is equal to the natural rate, and participation rates and average hours worked are at their trend levels, then LU = La / L, and equation (2) can be written in log form as

ln(Ya)=ln(A)+ln(U)+αln(K)+βln(La)(2)

Assuming that the rates of remuneration of labor and capital are equal to their marginal products and α and β are thus equal to their respective shares in factor income, equation (2’) can be estimated in the form

(ln(Ya)αln(K)βln(La))=ln(A)+ln(U)(2)

after A and U have been replaced by appropriate proxies such as time trends and indicators of utilization rates. Given the estimation results, the level of potential output can be derived by setting U and LU to one (at potential levels), which implies La = L, and A to the level predicted by the estimating equation, given U = LU = 1. 1/

  • The following variables were used to estimate equation (2’’): 2/

  • Ya: value added in the nonagricultural sector in 1990 constant prices;

  • K: stock of capital in the nonagricultural sector derived from a perpetual inventory model using data on total fixed investment in the nonagricultural sector and assuming an annual depreciation rate of 6 percent; 3/

  • La: total employment in the nonagricultural sector multiplied by an index of average hours worked;

  • L: potential labor input derived, as discussed in greater detail below, from population data using trend levels of participation rates and hours worked, and an estimate of the natural rate of unemployment;

  • α and β: income shares derived from the share of compensation of employees in domestic factor income (excluding agriculture); 1/ with the share of labor income rising by some 10 percentage points over the estimation period, different shares were used for sub-periods and the resulting series for labor and capital inputs were spliced; 2/

  • U: an indicator of the rate of factor utilization approximated by the ratio of actual to trend output in the nonagricultural sector; 3/

  • A: total factor productivity estimated by including trends for different sub-periods in the estimating equation.

In order to facilitate the assessment of the likely future contributions of labor input to potential output growth, L was decomposed into its main determinants as follows

L=LP·(γPRF+(1γ)PRM)·LNA·(100URN)·LH(3)

with LP indicating the total working age population, 4/ PRF and PRM the female and male participation rates in the total labor force, γ and (1-γ) the shares of females and males in the working age population, LNA the share of nonagricultural employment in the total labor force, URN an estimate of the natural rate of unemployment, and LH average hours worked. PRF, PRM, LNA, as well as LH are represented by three-year centered moving averages.

Historical developments in these variables are summarized in Chart V-1, which indicates that in addition to the growth of the working age population and some increase in the female labor force participation rate, the transfer of labor from the agricultural sector to the nonagricultural sector has played a major role in the growth of labor input in the latter sector. Thus, while the total working age population grew only at an average rate of 2.5 percent during 1972-95, employment in the nonagricultural sector increased at an average annual rate of 5.3 percent. With the steady decline in the share of agricultural employment in the labor force, however, the role of this factor has diminished in recent years. In addition, and presumably related to the inter-sectoral transfer of labor, the natural rate of unemployment in the nonagricultural sector appears to have fallen significantly (Chart V-2). 5/ The estimate of the natural rate of unemployment URN used to derive potential labor input takes possible links between these developments into account by relating the unemployment rate in the nonagricultural sector to the share of agricultural employment in the labor force. 1/

CHART V-1
CHART V-1

KOREA: DETERMINANTS OF LABOR INPUT IN THE NONAGRICULTURAL SECTOR, 1970–2005

Citation: IMF Staff Country Reports 1996, 136; 10.5089/9781451822021.002.A005

Source: Staff estimates.1/ Population 15 years and older; in millions.2/ Three-year centered moving averages.3/ Share of agricultural employment in total labor force.4/ Index 1971=100.
CHART V-2
CHART V-2

KOREA: UNEMPLOYMENT RATE, 1970–96

Percent

Citation: IMF Staff Country Reports 1996, 136; 10.5089/9781451822021.002.A005

Source: Staff estimates.1/ Seasonally adjusted.

Historical developments in investment ratios, capital stocks, and the capital output ratio are illustrated in Chart V-3. Associated with a sharp rise in the investment ratio 2/ in the second half of the 1970, from just over 15 percent in 1974 to well over 25 percent in 1979, the capital output ratio increased significantly. Following a decline during the 1980 recession, the investment ratio remained essentially flat for most of the 1980s but increased again in the early 1990s, with the capital output ratio rising from 2 in 1989 to 2.5 in 1995. These developments suggest that the relative growth contributions of capital, labor, and productivity have varied significantly in the past 25 years.

CHART V-3
CHART V-3

KOREA: INVESTMENT AND CAPITAL STOCKS, 1970–95

Percent

Citation: IMF Staff Country Reports 1996, 136; 10.5089/9781451822021.002.A005

Source: Bank of Korea; and staff estimates.1/ In 1990 prices; shares in total GDP.

Estimation of an equation similar to equation (2’’), with the dependent variable defined as the difference between actual value added in the nonagricultural sector and the weighted labor and capital inputs, yielded the following parameter estimates: 3/

article image

The variable t indicates a linear time trend for the whole estimation period, 1/ while t1, t2, and t3 cover the periods 1970-1978, 1979-1983, and 1991-1995, respectively, to allow for trend breaks. Finally, a1, a2, and a2 are dummy variables for the same periods reflecting shifts in the intercept.

The coefficients of the trend variables suggest significant changes in the growth of total factor productivity, which are clearly visible in Chart V-4. These shifts are discussed in detail in the next section. The output gaps implied by the estimates of potential output derived from equation (4) are shown in Chart V-5, together with the estimated unemployment gaps. 2/ Both variables are closely correlated but the variance of the output gaps is evidently much larger than that of the unemployment gaps, suggesting a very large Okun coefficient. 3/ In addition to some labor hoarding, the large Okun coefficient appears to reflect considerable cyclical variability in the labor force which has tended to limit cyclical changes in measured unemployment. 4/ The estimated output gaps for recent years suggest that the degree of resource pressures in the 1994-95 expansion was more moderate than in previous expansions. This result is consistent with the relatively subdued behavior of inflation.

CHART V-4
CHART V-4

KOREA: TOTAL FACTOR PRODUCTIVITY, 1972–95

Citation: IMF Staff Country Reports 1996, 136; 10.5089/9781451822021.002.A005

Source: Staff estimates.1/ Difference between value added in the nonagricultural sector and factor inputs.2/ Total factor productivity predicted by the equation described in the text after setting the proxy variable for syslical factors to zero and labor input to its normal level.
CHART V-5
CHART V-5

KOREA: OUTPUT GAPS, 1972–95

Percent

Citation: IMF Staff Country Reports 1996, 136; 10.5089/9781451822021.002.A005

Source: Staff estimates.1/ Difference between the log of actual output (seasonally adjusted) and the log of potential output, multiplied by 100.2/ Difference between trend unemployment rate (natural rate) and the actural unemployment rate (seasonally adjusted).

3. Past growth experience

The high growth rates registered in Korea and other East Asian economies, notably Hong Kong, Singapore, and Taiwan Province of China, in the past three decades have been the subject of an extensive debate in recent years. Sparked by the Word Bank’s study on the “East Asian Miracle,” which sought to draw the policy lessons from the success of these countries, 5/ the debate has increasingly focused on the nature of their remarkable growth performance. A key question has been whether growth in these countries was mainly extensive, driven by rapid increases in factor inputs, or intensive, driven by strong growth of total factor productivity (TFP). If the former was the case, it has been argued, their strong growth was perhaps not so difficult to understand, and, more importantly, was unlikely to be sustained in the future as it would be constrained by diminishing returns. This view in the future as it would be constrained by diminishing returns. This view has been expounded in particular by Krugman based mainly on empirical research by Young. 1/

The estimates of the growth contributions of capital, labor, and TFP derived from the analysis in the previous section seem, at first sight, to support this view (Table V-1). During 1972-1995, TFP is estimated to have grown at a rather modest annual rate of 1.7 percent, accounting for only 20 percent of potential output growth, with capital and labor each accounting for about half of the remaining 80 percent. These results are quite close to those obtained by Young although his study covers a somewhat different time period and is based on a relatively detailed breakdown of capital and labor inputs to allow for differences in the marginal products of the components. 2/

Table V-1.

Korea: Growth Contributions of Capital, Labor and Total Factor Productivity, 1972-95

(In percent; period averages)

article image
Sources: Staff estimates; and Young (1995).

For nonagricultural output.

The growth contributions were calculated from Young (1995), Table VII.

See “Korea--Recent Economic Developments” (SM/94/229, 8/29/94). These earlier estimates refer to total GDP, including agriculture, and are thus not fully comparable with the new estimates and those derived from Young (1995).

Previous staff estimates suggested an even smaller growth contribution of TFP and a much larger contribution of capital, reflecting capital stock data, which, as discussed above, probably exaggerated the growth of the capital stock. In addition, they were based on a relatively low average income share of labor derived from national accounts data that did not allow for the labor income of self employed and family workers. Finally, these estimates implied a relatively small growth contribution of labor, because they covered total GDP rather than value added in the nonagricultural sector and thus did not directly reflect the large shift of labor from the agricultural to the nonagricultural sector that has taken place in the past three decades. 3/

While the estimates for the whole period 1972-95 suggest that productivity growth has played a relatively modest role in Korea, there is evidence of a marked shift in the relative growth contributions of factor inputs and TFP between the 1970s and the 1980s. The extensive growth model seems to describe developments during the 1970s quite well, but around 1982-83, following the recovery from the recession of 1980, TFP growth picked up sharply, accounting for almost 40 percent of the growth of potential output. Factors such as human capital accumulation, which have been widely employed to explain differences in TFP growth in cross-country studies, 4/ do not seem to account for such a shift.

The change in the pattern of growth coincides with a change in policies from far-reaching government interventions in the allocation of resources associated with a drive to promote specific sectors, notably heavy and chemical industries, to a less interventionist approach which focused on general incentives for export-oriented activities. In addition, a gradual but comprehensive program of import liberalization was implemented, and interventions and controls in the financial sector were scaled back. 1/ This shift in structural policies, together with a successful program of macroeconomic stabilization, created an environment that was more conducive to an efficient use of resources than that of the 1970s, and may account for a large part of the change in the pattern of growth in the 1980s. It is doubtful that without this policy shift the high growth rates of the 1970s based largely on increases in factor inputs could have been sustained, let alone surpassed.

The strong growth of TFP appears to have continued until about 1990, followed by a slowdown to about 2 percent per year. This slowdown is more difficult to explain than the shift in the early 1980s. Reforms to deregulate the economy continued and were, in fact, accelerated in the first half of the 1990s. To some extent, the estimated slowing of productivity growth may be overstated due to some upward bias in the estimated increases in the capital stock since 1990, which in 1990-91 reflected partly a sharp increase in investment in residential buildings with a smaller marginal product than equipment and industrial structures. This bias could, however, explain at best part of the slowing of TFP growth. Thus, while productivity growth may well strengthen again in the coming years, an examination of future growth prospects needs to allow for the possibility that it may continue to grow at a considerably more moderate rate than in the 1980s.

4. Medium-term growth prospects: some scenarios

The purpose of the scenarios discussed in this section is to illustrate the implications for potential output growth over the medium term of alternative assumptions about developments in factor inputs and productivity. The projected growth rates are conditional on these assumptions and should thus not be interpreted as forecasts of a likely outcome.

The scenarios differ in the underlying assumptions regarding investment rates and growth of TFP. They have in common projections for the factors influencing labor input which imply a significant decline in the contribution of labor to potential output growth from 2.6 percent in 1995 to 1.6 percent by 2005. These projections are based on information about demographic and labor market developments, which suggest that a decline in the growth contribution of labor over the medium term is indeed likely, although the magnitude of the decline is subject to considerable uncertainty.

The projected slowing of the growth of labor input reflects the following factors. First, in view of the marked decline in the rate of population growth during the 1980s, from 1.7 percent in 1981 to less than 1 percent in 1990, the growth of the working age population is likely to slow further in the next decade. 1/ Second, the transfer of labor from the agricultural to the nonagricultural sector, which has played a key role in the past, is likely to be of minor importance in the future. Although the projections assume a further decline in the share of agricultural employment in the labor force, the contribution of this shift to the growth of the labor force in the nonagricultural sector is marginal, given the already low share of agricultural employment. Third, in view of developments in working hours in advanced industrial countries and in Korea during the 1980s, changes in average hours worked are more likely to have a negative than a positive impact on the growth of labor input in the next decade. The projections assume, nevertheless, unchanged average working hours. Fourth, with male labor force participation already at a high level, an increase in the overall participation rate will probably require a further rise in female participation. However, although the female participation rate is assumed to increase at a considerably faster rate than in the past, rising from less than 50 percent in 1995 to 60 percent in 2005, such an increase would offset only part of the slowdown in the growth of the working age population. 2/ Finally, given the low rate of unemployment, it is unlikely that the trend decline in the natural rate observed in the past will continue. 3/

With the growth contribution of labor likely to decline over the medium term, a key question concerns the conditions under which present growth rates of potential output can be maintained. This question is addressed in the scenarios summarized in Table V-2. Scenario 1 assumes, in addition to the developments discussed above in the factors influencing labor input, that the gross investment rate 4/ would remain broadly unchanged at the 1995 level, some 38 percent. TFP is assumed to continue to grow at about 2 percent as in recent years. The simulation results suggest that under these conditions it would be difficult to sustain growth at the rate of the recent past. With a growing share of gross investment needed to maintain the existing capital stock, the net investment ratio and the growth contribution of capital would decline over the medium term, in addition to the decline in the growth contribution of labor. As the wedge between gross and net investment increases with the rise in the capital output ratio, even a substantial increase in the gross investment ratio to 45 percent by 2005 (Scenario 2) would not prevent a slowing of potential growth over the medium term, although it would limit the extent of the decline.

Table V-2.

Korea: Potential Output Growth in 1996-2005--Alternative Scenarios

(In percent, period averages)

article image
Source: Staff estimates.

Nonagricultural sector.

Gross fixed investment in percent of GDP (both excluding agriculture).

Note: All scenarios assume the developments in labor input growth described in the text and a constant income share of labor of 0.7 over the medium term. In addition, they are based on the following assumptions:Scenario 1: an unchanged investment ratio from the 1995 level, with TFP continuing to growth at the 1995 rate.Scenario 2: a gradual increase in the investment ratio to 45 percent by 2005, with TFP continuing to grow at the 1995 rate.Scenario 3: an unchanged investment ratio from the 1995 level and a gradual increase in TFP growth to 3 percent by 2000.Scenario 4: an increase in the investment ratio to 40 percent by 1998 and a gradual increase in TFP growth to 3 percent by 2000.Scenario 5: an unchanged investment ratio from the 1995 level and a gradual increase in TFP growth to 3.4 percent by 2002.

These simulation results suggest that Korea’s medium-term growth potential depends critically on developments in the growth of TFP. If growth of factor productivity were to pick up gradually to about 3 percent by 2000, which would still be well below the estimated TFP growth for 1984-90, most of the decline in the growth contributions of labor and capital would be offset and potential growth could remain close to the present rate of 7 3/4 percent until about 2001, followed by a modest decline as the increase in TFP growth tapers off (Scenario 3). If in addition to these productivity gains the investment ratio were to rise to about 40 percent (Scenario 4), the present rate of potential output growth could be sustained until early in the next decade, but it would slow somewhat thereafter. Only a more pronounced increase in productivity growth to about 3.4 percent by 2002 would maintain potential growth at the present rate over the whole simulation period. Such productivity gains would mark a clear departure from the trend in the recent past. Korea’s past growth record suggests that they are not entirely infeasible, but it is debatable whether they are likely.

5. Conclusions

The estimates of potential output derived in this chapter on the basis of a simple production function approach appear to fit past developments quite well. The implied output gaps are broadly consistent with movements in the unemployment rate and recent price developments. In addition, the estimates provide interesting insight into the sources of the rapid growth registered in the past two and a half decades. In the 1970s, the expansion of the Korean economy appears to have been attributable mainly to large increases in factor inputs, both capital and labor, while growth of total factor productivity played a negligible role. The picture changed significantly in the 1980s, with a strong pickup in productivity growth more than offsetting declining growth contributions of factor inputs. With recent demographic trends pointing to a steady slowing of the growth of the working age population over the medium term, the decline in the growth contribution of labor, evident already in the past two decades, is likely to continue. The medium-term growth scenarios discussed in this chapter suggest that even a relatively large increase in the investment ratio would only offset a small part of the resulting decline in potential output growth. Without a pickup of total factor productivity growth following the moderation in recent years, Korea’s potential growth rate could decline by as much as 1 1/2 percentage points over the medium term.

References

  • Adams, Charles and David T. Coe, 1992, “A System’s Approach to Estimating the Natural Rate of Unemployment and Potential Output for the United States,” in The Unites States Economy, Performance and Issues, ed. by Y. Horiguchi and others, (Washington: International Monetary Fund) pp. 252-313.

    • Search Google Scholar
    • Export Citation
  • Adams, Charles, Paul R. Fenton and Flemming Larsen, 1987, “Potential Output in Major Industrial Countries,” IMF Staff Studies for the World Economic Outlook (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Bank of Korea, National Accounts (Seoul, various issues).

  • Cho, Yoon-Je and David C. Cole, 1992, “The Role of the Financial Sector in Korea’s Structural Adjustment,Structural Adjustment in a Newly Industrialized Country: The Korean Experience, ed. by V. Corbo and Sang-Mo Suh, (Baltimore and London: Johns Hopkins University Press for the World Bank), pp. 115-37.

    • Search Google Scholar
    • Export Citation
  • Clark, Peter, Douglas Laxton, and David Rose, 1996, “Asymmetry in the U.S. Output in Inflation Nexus,IMF Staff Papers, Volume 43 (1), (Washington: International Monetary Fund), pp. 216-51.

    • Search Google Scholar
    • Export Citation
  • Coe, David T. and C. John McDermott, 1996, “Does the Gap Model Work in Asia?IMF Working Paper (WP/96/69), (unpublished; Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Fukuda, Shin-ichi and Hideki Toya, 1994, “The Role of Human Capital Accumulation for Economic Growth in East Asian Countries,Seoul Journal of Economics, Volume 7 (2), pp. 213-31.

    • Search Google Scholar
    • Export Citation
  • Hodrick, Robert and Edward Prescott, 1980, “Post-War U.S. Business Cycles: An Empirical Investigation,Carnegie-Mellon University Discussion Paper No. 451 (Pittsburgh: Carnegie-Mellon University).

    • Search Google Scholar
    • Export Citation
  • Kim, Jong-Il and Lawrence Lau, 1994, “The Sources of Economic Growth of the East Asian Newly Industrialized Countries,Journal of Japanese and International Economies, Volume 8, pp. 235-71.

    • Search Google Scholar
    • Export Citation
  • Krugman, Paul, 1994, “The Myth of Asia’s Miracle,Foreign Affairs, Volume 73 (November-December), pp. 62-78.

  • Lucas, Robert E., 1993, “Making a Miracle,Econometrica, Volume 61 (2) pp. 251-72.

  • National Bureau of Statistics, Republic of Korea, Annual Report on the Economically Active Population Survey (Seoul, various issues).

  • Page, John, 1994, “The East Asian Miracle: Four Lessons for Development,NBER Macroeconomics Annual, pp. 219-281.

  • Prescott, Edward C., 1986, “Theory Ahead of Business Cycle Measurement,Carnegie-Rochester Conference Series on Public Policy, No. 25, pp. 11-44.

    • Search Google Scholar
    • Export Citation
  • Pyo, Hak-kil, 1988, “Estimates of Capital Stocks and Capital Output Coefficients by Industries for the Republic of Korea (1953-1986),Korea Development Institute Working Paper, No. 8810 (Seoul).

    • Search Google Scholar
    • Export Citation
  • Sarel, Michael, 1995, “Growth in East Asia: What We Can and What We Cannot Infer From ItIMF Working Paper (WP/95/98) (unpublished; Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • World Bank, 1993, “The East Asian Miracle: Economic Growth and Public Policy,” (New York: Oxford University Press).

  • Young, Alwyn, 1995, “The Tyranny of Numbers: Confronting the Statistical Realities of the East Asian Growth ExperienceQuarterly Journal of Economics, Volume 110 (3), pp. 641-80.

    • Search Google Scholar
    • Export Citation
  • Young, Soo-Gil, 1992, “Import Liberalization and Industrial AdjustmentStructural Adjustment in a Newly Industrialized Country: The Korean Experience, ed. by V. Corbo and Sang-Moh Suh, (Baltimore and London: Johns Hopkins University Press for the World Bank) pp. 171-203.

    • Search Google Scholar
    • Export Citation
1/

This chapter was prepared by Ms. Marianne Schulze-Ghattas (ext. 37618) who is available to answer questions.

2/

An alternative view, associated with theories of real business cycles, is that fluctuations in output growth reflect mainly changes in supply conditions. See, for example, Prescott (1986).

3/

See, for example, World Bank (1993), Page (1994), Krugman (1994), and Young (1995). For a reexamination of the debate see Sarel (1995).

4/

Previous estimates are discussed in SM/94/229, 8/29/94.

1/

For a brief summary of alternative approaches to the estimation of potential output see, for example, Adams, Fenton, and Larsen (1987).

2/

Many studies involving estimates of potential output have used Hodrick-Prescott filters to generate the estimates (see, for example, Coe and McDermott (1996)). This method is described in Hodrick and Prescott (1980).

1/

This approach is a slightly modified version of the framework used in Adams, Fenton, and Larsen (1987).

2/

Data on GDP, sectoral value added, and investment were taken from: Bank of Korea, “National Accounts,” various issues; data on employment, labor force, working age population, and hours worked from: National Bureau of Statistics, Republic of Korea, “Annual Report on the Economically Active Population Survey,” various issues. All series were seasonally adjusted.

3/

Previous staff estimates of potential output for Korea (see SM/94/229, 8/29/94) used estimates of the capital stock provided in Pyo (1988). These estimates combine investment data from the national accounts and information from wealth surveys, which are carried out at intervals of ten years. For many component series, the cumulated investment between two wealth surveys was smaller than the change in stocks implied by the surveys, a problem also discussed in Young (1995). Pyo combined data from the two sources on the assumption that both are subject to error. By contrast, the estimates used in this study assume that the investment data from the national accounts are more reliable. The capital stock series were initialized by assuming a capital output ratio of 1.5 in the starting year 1953. The data for the estimation period (1970 onward) are, however, not very sensitive to changes in the starting value.

1/

The national income accounts data on compensation of employees were adjusted to include the imputed labor income of self employed and family workers, based on employment survey data for the latter.

2/

The following income shares for labor were used: 0.65 (1970-1980), 0.67 (1981-1989), 0.70 (1990-1995).

3/

Trend output was estimated using a Hodrick-Prescott filter.

4/

Total population 15 years and older.

5/

The large intersectoral transfer of labor was probably associated with a relatively high level of frictional unemployment.

1/

Ideally, the natural rate of unemployment and potential output should be estimated jointly in a Phillips-curve-type framework (see, for example, Adams and Coe (1992)). In this study, the variable used to approximate the natural rate was derived from the predicted values from the following equation for the unemployment rate UR (t values in parentheses)

UR =0.64+0.16 LAGR2(adj.):0.90
(1.79)(14.34)DW:0.93

with LAG indicating the share of agricultural employment in the labor force.

2/

For the nonagricultural sector.

3/

The equation was estimated with quarterly data for the period 1970 to 1995.

4/

A Cochrane-Orcutt adjustment was used to account for serially correlated errors. The estimated first-order autocorrelation coefficient was 0.84.

1/

The variable takes the value 1970 in the first quarter of 1970 and increases in increments of 0.25.

2/

The output gap as a percentage of potential output is approximated by the difference between the log of actual output and the log of estimated potential output, multiplied by 100. The unemployment gap is defined as the difference between the estimated natural rate of unemployment and the actual unemployment rate.

3/

The estimated Okun coefficient for the period 1970-1995 is 11, a multiple of the coefficient typically observed in countries such as the United States.

4/

The absence of an unemployment insurance scheme, which was introduced only in mid 1995, may explain part of the cyclical variability of the labor force in Korea.

2/

Kim and Lau (1994), based on the estimation of a “meta-production function” using intercountry data, also conclude that technical progress has played at best a small role in the growth experience of the newly industrialized East Asian countries.

3/

If the growth contributions of capital, labor, and TFP are estimated for total GDP in economies undergoing rapid change in the sectoral composition of value added, the effects of the structural shifts should be reflected in a close relationship between TFP growth and indicators of the composition of GDP.

4/

See, for example, Fukuda and Toya (1994).

1/

For a description of the policy changes in the early 1980s see Young (1992) as well as Cho and Cole (1992).

1/

The projections assume a decline in the growth of the working age population (population 15 years and older) from 1.7 percent in 1996 to 1 percent in 2005, broadly in line with developments in total population growth 15 years earlier, assuming constant death rates.

2/

The estimates indicate that without an increase in the participation rate, the growth contribution of labor would fall to less than 1.5 percent by the middle of the next decade.

3/

The projections assume an unchanged natural rate of unemployment.

4/

Gross fixed investment in the nonagricultural sector relative to GDP in that sector.

Korea: Selected Issues
Author: International Monetary Fund