1. Since the early 1980s, the current account of the balance of payments has improved substantially, shifting from a deficit of ½ percent of GDP in 1980, to a surplus of about 5¾ percent in 1996 (Table 1 and Figure 1). The current account surplus is now one of the highest in the world—although statistical problems may result in an overestimation of the surplus (see Box 1). This section examines the causes of this increase from both a balance of payments and a saving-investment perspective. It first considers each perspective separately, and then discusses the consistency of the two approaches. In conclusion, it addresses the question how the underlying current account compares with the medium-term trend in the saving-investment balance; or, in other words, whether there is any indication of policy tensions, and how such tensions might be resolved under EMU.

Abstract

1. Since the early 1980s, the current account of the balance of payments has improved substantially, shifting from a deficit of ½ percent of GDP in 1980, to a surplus of about 5¾ percent in 1996 (Table 1 and Figure 1). The current account surplus is now one of the highest in the world—although statistical problems may result in an overestimation of the surplus (see Box 1). This section examines the causes of this increase from both a balance of payments and a saving-investment perspective. It first considers each perspective separately, and then discusses the consistency of the two approaches. In conclusion, it addresses the question how the underlying current account compares with the medium-term trend in the saving-investment balance; or, in other words, whether there is any indication of policy tensions, and how such tensions might be resolved under EMU.

I. the current account of the netherlands1

A. Introduction

1. Since the early 1980s, the current account of the balance of payments has improved substantially, shifting from a deficit of ½ percent of GDP in 1980, to a surplus of about 5¾ percent in 1996 (Table 1 and Figure 1). The current account surplus is now one of the highest in the world—although statistical problems may result in an overestimation of the surplus (see Box 1). This section examines the causes of this increase from both a balance of payments and a saving-investment perspective. It first considers each perspective separately, and then discusses the consistency of the two approaches. In conclusion, it addresses the question how the underlying current account compares with the medium-term trend in the saving-investment balance; or, in other words, whether there is any indication of policy tensions, and how such tensions might be resolved under EMU.

Table 1.

Netherlands: Balance of Payments

(In percent of GDP)

article image
Sources: IMF, Balance of Payments Statistics; and data provided by the authorities.
FIGURE 1
FIGURE 1

NETHERLANDS: Current Account and Its Components

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: IMF, Balance of Payments Statistics.

B. The Current Account from a Balance of Payments Perspective

2. A breakdown of the current account into its various components reveals that the striking improvement in the current account can be attributed mainly to the improvement of the trade balance. The other components of the current account have either not shown a clear trend, or quantitatively were not very important.

3. Since the early 1980s, the trade balance has improved by about 5½ percent of GDP (Table 1 and Figure 1). This improvement came in two phases. First, between 1980 and 1982 the balance switched from a deficit of 0.1 percent to a surplus of 4½ percent of GDP. Having fluctuated around this level for about a decade, the trade surplus then increased further to about 5½ percent of GDP in 1993. The contribution of other components of the current account was not as important. The income balance has registered no clear trend, which is somewhat surprising in view of the large current account surpluses in the past decade and a half (see below). The services balance did not change much during the 1980s, but increased by about 1 percentage point during the 1990s. This was offset by the balance on current transfers, which has seen a gradual deterioration, and is now 1 percentage point lower than in the early 1980s.

Exchange rate developments and competitiveness

4. Between the mid-1970s and the mid-1980s, Dutch competitiveness improved considerably (Figure 2). Measured by the unit labor cost-based real effective exchange rate, the real effective exchange rate declined by 25 percent between 1977 and 1984. This improvement (which reversed, but not fully, a deterioration of competitiveness in the 1960s and 1970s) resulted from an impressive performance of unit labor costs; the trade-weighted nominal effective exchange rate continued to appreciate—although there was some depreciation vis-à-vis the deutsche mark, and a strong depreciation vis-à-vis the U.S. dollar.

Is the Current Account Surplus Really So High?

There are two reasons why official balance of payments statistics may overestimate the current account surplus:

  • First, there is a substantial difference between the current account on a cash basis and on a transactions basis. In 1996, for instance, the current account on a transactions basis amounted to 5.7 percent of GDP, while the current account on a cash basis was only 3.1 percent of GDP. The discrepancy is mainly due to the trade balance, with goods imports on a cash basis far higher than goods imports on a transaction basis.

  • Second, the current account does not capture revaluations of foreign investments. In the past decade and a half there have been substantial revaluation losses on Dutch investment abroad, and 15 years of current account surpluses have been associated with a decline in the net foreign investment position (see below).

In addition, it is important to note that the current account measure captures the retained earnings of foreign subsidiaries of Dutch multinationals. These earnings are booked on the current account as investment income, and then registered on the financial account as an outflow of direct investment. Thus, while they enter the Dutch current account surplus, they are not generated in the Dutch economy. In 1995, reinvested earnings amounted to some 1.3 percent of GDP. Technically, this is internationally accepted as the correct classification of such flows, but it reflects developments outside the geographic confines of the Dutch economy.

While these factors may affect the level of the current account surplus, they are less likely to affect trends—although the discrepancy between the trade balance on a cash basis and a transactions basis has been growing over time.

FIGURE 2
FIGURE 2

NETHERLANDS Exchange Rate Indicators

(1977=100)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: IMF, IFS; and OECD, Analytical Database.

5. The real exchange rate depreciation was associated with a substantial improvement in profitability of enterprises: the share of labor income in value added declined dramatically (Figure 3). Thus, the depreciation was not only used to gain market share but also to restore profit margins (see Box 2, Appendix II, and the discussion in the next section).

FIGURE 3
FIGURE 3

NETHERLANDS REER and Labor Income Share

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: IMF, World Economic Outlook.

6. Between 1985 and 1995, the real effective exchange rate did not change much, as a favorable performance in unit labor costs was offset by an appreciation of the nominal effective exchange rate. Since 1995, however, there has been a renewed depreciation of the real effective exchange rate. This was mainly due to a change in the nominal effective exchange rate—resulting from a depreciation of the guilder vis-à-vis the U.S. dollar, the pound sterling, and the lira.

Domestic demand

7. Developments in domestic demand have had a pronounced cyclical impact on the trade balance. The sharp increase in the Dutch trade balance in the 1980–82 period was associated with a decline in imports, rather than a surge in exports, and this decline in imports was associated with very weak domestic demand (Figure 4). The increase in the trade balance in 1993 also was associated with very weak domestic demand. In both periods business investment, which has a very high import component, declined strongly. As the Netherlands exports relatively few investment goods, and instead is specialized in products such as food and food products (Table 2), exports generally decline less than imports during European-wide recessions, resulting, ceteris paribus, in an improvement in the trade balance in such periods. However, an assessment of the role of domestic demand in the structural increase of the current account—rather than its cyclical role—is only possible from a saving-investment perspective, and this is discussed in Section C below.

FIGURE 4
FIGURE 4

NETHERLANDS Domestic Demand and Its Components

(1980=100)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: OECD, Economic Outlook Database.
Table 2.

Netherlands: Trade Balance and Its Components

(In percent of GDP)

article image
Source: UNCTAD.

The income balance and the net foreign investment position

8. With large current account surpluses since the early 1980s, the net foreign investment position of the Netherlands would have been expected to increase strongly, and to have triggered, in turn, a strong increase in the income balance. However, the net foreign investment position has actually decreased, from 32¾ percent of GDP in 1985, to 13½ percent in 1995 (Table 3).

Table 3.

Netherlands: Net Foreign Investment Position

(In percent of GDP)

article image
Source: IMF, Balance of Payments Statistics.

9. According to research by the Central Planning Bureau, several factors account for this surprising development:

  • An important factor has been the strong increase in stock prices since the early 1980s. An increase in stock prices in world markets leads (ceteris paribus) to a deterioration of the Dutch net foreign investment position, because foreign holdings of Dutch shares exceed Dutch holdings of foreign shares.2 Thus, if both assets and liabilities increase by the same percentage, the net investment position of the Netherlands will deteriorate. The deterioration has been exacerbated by the fact that the increase in Dutch stock prices in the 1990s was stronger than that in many other countries.

  • Another contributing factor has been the appreciation of the guilder. As a large portion of Dutch foreign assets is denominated in foreign currencies, an appreciation of the guilder will lead to a reduction of the net foreign investment position in Dutch guilders.

  • A third factor is the underestimation of Dutch direct investment abroad Foreign direct investment is typically valued at historical prices rather than market prices. This underestimates the value of direct investment, especially in countries with high inflation and a depreciating currency.

  • Finally, the return on foreign direct investment in the Netherlands has been well above the return on Dutch foreign investment abroad. The difference is in large part due to the United States: the return on U.S. investment in the Netherlands has been much higher than the return on Dutch investment in the United States. While it is not entirely clear what lies behind this difference, a U.S. Department of Commerce study found that it is not unique to the Netherlands.

Pricing in Export Markets 1/

The extent to which changes in cost competitiveness are transmitted to changes in relative export prices depends on pricing behavior in international markets. If exporters folly pass on their cost changes, export prices will move in line with cost competitiveness; if they folly absorb cost changes in profit margins, cost competitiveness and export prices will be unrelated:

Px/Px*=δ+βCx/Cx*(1)

where Px/Px* is the relative export price in domestic currency, and Cx/Cx* is the relative cost term. If β = 1, exporters will folly reflect changes in cost competitiveness into their prices, leaving profit margins unchanged.

Estimation of equation (1) for the Dutch manufacturing sector suggests that only 40 percent of change in relative labor costs is reflected in export prices. In other words, 60 percent of relative labor costs changes are not reflected in export prices, which implies that an improvement in cost competitiveness has a substantial impact on profit margins.

1/ This box and the related Appendix I were prepared by Meral Karasulu.

C. The Current Account from a Saving-Investment Perspective

10. From a saving-investment perspective, the increase in the current account since the early 1980s mainly reflects an increase in the saving surplus of enterprises (Figures 5 and 6). The salving surplus (i.e., saving minus investment) of households has not shown a clear trend, while the saving surplus of government, after having decreased strongly in the 1970s, did not show a sizable change in the 1980s and 1990s.

FIGURE 5
FIGURE 5

NETHERLANDS National Saving Surplus and Its Components

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: IMF, World Economic Outlook.
FIGURE 6
FIGURE 6

NETHERLANDS Saving Surpluses of the Various Sectors

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: IMF, World Economic Outlook.

11. The increase in the various periods can be broken down as follows:

  • Between 1980 and 1984, the saving surplus increased by 5½ percentage points, from a deficit of 1¼ percent, to a surplus of 4¼ percent. This improvement was mainly associated with an increase in the saving surplus of enterprises; the saving surplus of the government deteriorated somewhat (although the government deficit declined—see below), while the saving surplus of households increased only very little.

  • Between 1985 and 1992, the saving surplus declined by 1¼ percentage points. The saving surplus of enterprises declined by 2½ percentage points, which was only partly offset by an increase in the saving surplus of households.

  • From 1992 onwards, the saving surplus increased further, from some 3 percent of GDP in 1992, to some 5 percent in 1993, and 5¾ percent in 1997. This increase was mainly the result of an increase of the saving surplus of enterprises, although toward the end of the period, the saving surplus of government increased also.

Enterprise saving surplus

12. The increase in the saving surplus of enterprises in the early 1980s was substantial. Between 1980 and 1984, the saving surplus rose by more than 6 percentage points, from a deficit of 1¼ percent of GDP to a surplus close to 5 percent. Between 1980 and 1982, this increase was mainly associated with a decline of investment; but in the following 2 years it reflected a strong recovery in saving, which outpaced the increase in investment. Since the early 1980s, the saving surplus of enterprises has displayed a cyclical pattern, but it has been at a structurally higher level than in the 1970s. Several exogenous factors could be behind this shift:

  • Real long-term interest rates rose sharply in the late 1970s. Following their fluctuation around zero percent in the 1960s and early 1970s, they increased to over 6 percent in the early 1980s, and did not fall below 5 percent until 1996.

  • Demand growth was slower during a large part of the 1980s, which reduced the need for capacity expanding investment. Wage growth was also much less rapid which slowed down the substitution of capital for labor by capital deepening investment.

  • There was a large decline in the labor income share, which boosted corporate profits and hence savings. (Of course, it also stimulated investment by making it more profitable.)

13. The shift may also have reflected a change in the risk preference of enterprises. The large bank borrowings of the 1970s made firms very vulnerable in the recession of the early 1980s, when demand declined sharply, while interest rates rose for a protracted period. This experience may have increased firms’ risk perception of bank loans and reduced their willingness to finance investment through loans—a pattern that has been observed in Belgium also. Thus, firms now finance their investments almost entirely out of retained earnings. However, empirical analysis by the staff did not detect a clear change in enterprise behavior: in the 1980s and 1990s, corporate investment and corporate saving appear to have behaved in broadly the same way as in the 1970s (see Appendix I):

  • The ratio of corporate investment to GDP tracks real GDP growth, with rapid GDP growth associated with a high investment ratio (Figure 7). Thus, the slowdown of investment during the 1970s, and its collapse during the early 1980s, mirrored developments in GDP growth—which in turn reflected developments in world trade growth (Figure 8), as did the recovery of investment in the second half of the 1980s.

  • The ratio of corporate saving to GDP closely follows the capital income share (Figure 9).3 Thus, the boost in corporate saving in the first half of the 1980s was associated with a strong increase in the capital income share.

FIGURE 7
FIGURE 7

NETHERLANDS Corporate Investment and Real GDP Growth

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

FIGURE 8
FIGURE 8

NETHERLANDS Export Market Growth and GDP Growth

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

FIGURE 9
FIGURE 9

NETHERLANDS Capital Income Share and Corporate Saving

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

14. Thus, the most important factor behind the structural shift in the saving surplus would seem to be the structural decline in the labor income share.4 The cyclical decline in investment in the early 1980s also boosted the surplus, but only on a temporary basis. Other factors, such as the increase in real interest rates, and a more cautious behavior of enterprises, may also have played a role, but their role is difficult to demonstrate empirically.

15. An important reason for the high level of the enterprise saving surplus may be the dominating position in the Dutch economy of multinational firms. Albeit it a small country, the Netherlands harbors quite a few very large multinational firms (Royal Dutch Shell, Unilever, Phillips, and Elsevier Reed). As these firms expand worldwide, rather than in the domestic market only, they invest a large part of their profits abroad, resulting in the combination of a large corporate saving surplus and substantial foreign direct investment outflows. Thus, in 1996, for instance, a corporate saving surplus of 4.4 percent of GDP was combined with a net foreign direct investment outflow of 3.9 percent of GDP.

Government saving surplus

16. The government saving surplus, which deteriorated sharply during the 1973–82 period, has improved by about 2 percentage points since 1982. It may seem surprising that the continuous fiscal consolidation since 1982—which resulted in a cut in the expenditure ratio of more than 13 percentage points—has not been associated with a greater increase in the saving surplus. However, about 40 percent of the expenditure cut reflected a cut in on-lending and capital transfers—both of which do not affect the saving surplus; and the remaining expenditure cuts were to a large extent offset by a reduction of the tax burden (Tables 4 and 5, and Figure 10):

Table 4.

Netherlands: General Government Finances, 1982–96

(In percent of GDP, on cash basis)

article image
Source: Data provided by the authorities.
Table 5.

Netherlands: General Government Expenditure Reduction and Increase in Saving Surplus

(In percent of GDP)

article image
FIGURE 10
FIGURE 10

NETHERLANDS Government Saving Surplus and Net Lending

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: IMF, World Economic Outlook.
  • About 30 percent of the expenditure decline reflected a cut in on-lending for social purposes, especially to housing corporations. The cut in on-lending did slow the growth of gross debt—but it changed neither the EMU-deficit, nor the saving surplus.

  • Another 10 percent of the expenditure cuts reflected a cut in capital transfers, especially to enterprises. A cut in capital transfers does reduce the EMU-deficit, but it does not change the saving surplus.

  • A large part of the other expenditure cuts were offset by declines in revenues. The share of tax revenue in GDP declined by almost 3 percentage points, as taxes were cut to stimulate wage moderation, and to improve the functioning of labor markets. Nontax revenues declined as well, especially after the collapse of oil prices in 1986, as a result of which natural gas revenues dropped from 4¾ percent of GDP in 1986, to 1¾ percent in 1988.

Household saving surplus

17. The household saving surplus has tended to vary over the economic cycle, but the pattern during recessions has evolved in an interesting manner. During the economic crisis of the early 1980s, the saving surplus behaved procyclically: as household confidence declined, when the economy went into recession, the saving surplus increased. During the economic slowdown in the early 1990s, by contrast, saving behaved counter-cyclically: with consumer confidence relatively strong, households reduced their saving surplus to smooth consumption when the economy slowed down.

18. The household saving surplus has shown, however, no clear structural trend over the past three decades. Both saving and investment are at about the same level relative to GDP as they were in the early 1970s. However, while total household saving has not changed much, its composition has changed substantially. The share of saving that is channeled through pension funds and life insurance companies (“contractual” saving) has increased, while the part that is channeled through banks or other institutions (“free” saving) has gone down.

19. The increase in contractual saving mainly reflects a surge in investment income of pension funds, associated with a maturing of fully funded pension plans (Figure 11). Most pension plans were founded in the fifties and sixties, and as the amount of built-up capital increased, so did investment income. There has probably also been some substitution of contractual for free saving, especially life insurances and annuities. Tax treatment of contractual saving is far more favorable, and while the tax regime has not been changed much in favor of contractual saving, the awareness of consumers of the tax advantages has increased, in part due to aggressive marketing by insurance companies.5

FIGURE 11
FIGURE 11

NETHERLANDS Components of Household Saving

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 102; 10.5089/9781451829327.002.A001

Source: CBS, National Accounts.

20. It is likely that saving in the Netherlands is stimulated by the fact that the pension system is funded. In other words, if pensions had been wholly financed on a pay-as-you-go basis, personal saving would likely to have been somewhat lower (see Box 3). The magnitude of this effect is hard to gauge, but if the pension system had not been funded, saving might plausibly have been some 1 to 2 percentage points of GDP lower.

D. The Link Between Both Approaches and the Current Account in the Future

21. The analyses in the previous two sections are, of course, complementary. From a balance of payments perspective, the main factor behind the structural increase in the current account emerges as the exchange rate depreciation (while cyclical increases in the current account appear to have been triggered by weak domestic demand). From a saving-investment perspective, on the other hand, the dominant factor behind the structural increase appears to have been the sharp increase in the corporate saving surplus. A key element underlying the increase in both the saving-investment balance and the export-import balance was the decline in the labor income share. By shifting income to a sector with a high marginal saving rate, the decline in the labor income share strengthened the saving-investment balance; and by reducing domestic demand and hence imports, it improved the export-import balance.

22. Of course the strong improvement in the current account in the early 1980s was to a large extent a reversal of developments in the 1960s and 1970s, when competitiveness had deteriorated strongly. As noted earlier, while a real exchange depreciation of 25 percent between 1977 and 1984 is very large when seen in isolation, it essentially brought competitiveness back to the level of the mid-1960s.

23. Seen in this much longer time horizon, the high level of the current account seems to reflect several underlying factors: a relatively young population in the face of a demographic transition that will be more severe than in many other countries; a pension system that is funded; a relatively high per capita income; and the presence of a number of very large multinationals that invest a substantial part of their earnings abroad.

Does the Funding of the Pension System Increase Household Saving?

Does the fact that a large part of the Dutch pension system is funded augment saving? In other words, would saving have been lower if there had been a pay-as-you-go rather than a partially funded system?

Mechanically, one might expect that for given pension benefits, saving by workers themselves will not depend on whether pensions are financed through a folly funded system or a pay-as-you-go system. Thus, the fact that a pension fond sets aside money, while benefits remain unchanged, will not alter the amount set aside by workers themselves—and every guilder saved by pension funds will add to saving.

However, there are several reasons why a guilder set aside by pension funds will in practice not increase saving by the foil amount. First, in a funded pension system it may be more credible that promised pensions will indeed be paid. Thus, precautionary saving may be lower under a funded pension system. Second, in a funded system, pension premiums of future workers will be lower, and their lifetime income will be higher. To the extent that current workers behave as if they have an infinite planning horizon, and take into account utility of future generations as well, they will lower their bequests—and hence their savings—under a funded pension system.

The discussion is very similar to the Ricardian debt equivalence debate that was reignited by Barro (1974). According to Barro, it does not matter whether government expenditure is financed through taxes or through bonds, as any increase in government debt will be offset by higher bequests to future generations. In other words, changes in government saving are fully offset by changes in private saving; and total saving does not depend on government saving. In the same vein, it could be argued that it does not matter whether pensions are funded or paid on a pay-as-you-go basis: any increase in saving through pension funds will be offset by a reduction in other forms of saving.

While foil debt equivalence is generally not found to hold, empirically there seems to be some degree of equivalence, with, as a general proposition, each additional unit in government saving being offset by a decline in private saving of some 50 percent. The same argument could be applied to pensions.

Overall, it would thus seem plausible that saving will be higher under a funded pension system. However, as other forms of personal saving are likely to be somewhat lower under a funded pension system, the increase is not one-for-one. As an illustration, if every guilder saved by a pension fond increased saving by 30 cents, this would imply that, with saving through pension schemes amounting to some 5 percent of GDP (the remainder of contractual saving relating to life insurances), the effect on total saving would be about 1¼ percent of GDP, while if it increased saving by 50 cents, the effect would be 2½ percent of GDP.

24. Against this background, the question can be considered what these two approaches to the current account suggest for developments in the near future:

  • Looking at the saving-investment balance, one would not expect to see much current account improvement. Corporate profits will be boosted by the lagged effect of exchange rate depreciation; but with the economy at potential, this to a considerable extent will be offset by an increase in private investment. Fiscal consolidation can also be expected to continue, but with a deficit that is already small, the increase in the saving-investment balance is unlikely to be large.

  • From a balance of payments perspective, one would expect some further increase in the surplus, as the present level is somewhat below the underlying current account level Both the lagged impact of the real effective exchange rate depreciation since 1995 and the closing of the output gaps in trading partners would point to an increase in the current account balance. On the other hand, with the domestic economy close to potential currently, it is not clear whether enterprises will develop capacity over the near to medium term to meet such additional demand fully, while pressures on resources could trigger wage and price effects that would result in a deterioration in competitiveness, thus reducing the underlying current account balance. The actual improvement in the current account could thus be less than indicated by a projection based on the present real exchange rate and the closing of foreign output gaps.

25. The above analysis is indeed broadly borne out by a multilateral assessment of trends in current account balances.6 Starting from a 1997 current account surplus of slightly less than 6 percent, this suggests that the underlying current account balance is some 8-9 percent of GDP—somewhat higher than a long-run trend in the saving investment balance of some 6-7 percent of GDP. This suggests that the economy enjoys a high degree of competitiveness, partly perhaps in reflection of a cyclical weakness of the core ERM currencies against the U.S. dollar. Over time, the tensions implied by these estimates could well be resolved by a modest degree of real appreciation (through price and wage effects, or a reversal of earlier nominal depreciation of the currencies in the core ERM/future euro). At the same time, however, it is likely that there will be somewhat higher rates of investment in the Netherlands: to the extent this higher investment responds to structural reforms, it would tend to dampen any increase in the medium-term saving-investment balance.

APPENDIX I: HAS THERE BEEN A BREAK IN ENTERPRISE BEHAVIOR?

26. This appendix tries to establish whether there has been a break in enterprise behavior in the early 1980s. It will first look at corporate investment and corporate saving separately, and then at the corporate saving balance.

Corporate Investment

27. Several variables can be expected to affect the ratio of corporate investment to GDP: demand growth, profitability, and real interest rates. However, while several indicators for profitability, real interest rates, and demand growth were examined, only demand growth variables were found to have a statistically significant impact. Table 6 shows the regression results for the final specification. In this specification, the investment rate depends upon the growth rate of real GDP: the faster the growth of real GDP, the higher the investment rate.

Table 6.

Netherlands: Regression Results: Corporate Investment

article image
Source: Fund staff calculations.Note:t-values are in parentheses.Variables are:iy Corporate investment as percent of GDPsy Corporate saving as percent of GDPgrgdp Growth of real gdp (average of current and previous year)grworld Growth of export markets (average of current and previous year)

28. The regression results in Table 6 do not indicate that there has been a break in investment behavior. Indeed, the coefficients in the equation estimated for the 1970–80 period are statistically indistinguishable of those for the 1980–96 or 1983–96 period.

29. While theoretically the link between high growth and high investment could go in both directions, it is likely that the causality has mainly run from high growth to high investment. If the investment equation is re-estimated using growth of relevant world markets rather than growth of GDP, the fit of the equation remains virtually unchanged. Indeed, as indicated in Table 6, a strong correlation between growth of world markets and GDP growth can be found.

Corporate Saving

30. With little fluctuation in corporate dividends, corporate saving can be expected to mainly depend on corporate profits.7 The capital income share, which is defined as that part of value added that does not go to labor, serves as a proxy, although not a perfect one, of corporate profits. Thus, we would expect to find a strong relation between the capital income share and corporate savings.

31. As indicated in Table 7, such a relationship can indeed be found. Fluctuations in the capital income share explain a very large part of fluctuations in corporate saving. Further econometric tests did not reveal any break in corporate saving behavior in the 1980s.

Table 7.

Netherlands: Regression Results: Corporate Saving

article image
Source: Fund staff calculations.Note:t-values are in parentheses.Variables are:sy Corporate saving as percent of GDPcsh Share of capital income in value added

Corporate Saving Surplus

32. With both corporate saving and corporate investment displaying stable behavior in the past few decades, we would expect to find a stable relationship for the corporate saving balance as well. In such a relationship, we would expect the following variables to be statistically significant: the growth rate of real GDP and the share of capital income in value added. As indicated in Table 8, such a relationship can indeed be found.

Table 8.

Netherlands: Regression Results: Corporate Saving Surplus

article image
Source: Fund staff calculations.Note:t-values are in parentheses.Variables are: bal Corporate saving surplusgrgdp Growth of real gdp (average of current and previous year)csh Share of capital income in value added

Conclusion

33. There is no indication that there has been a break in corporate behavior in the early 1980s. Indeed, for investment, saving, as well the saving surplus a specification could be found that showed a stable behavior throughout the period.

APPENDIX II: AN EMPIRICAL ANALYSIS OF THE EFFECT OF EXCHANGE RATE CHANGES ON DUTCH EXPORTS AND IMPORTS8

34. This section will investigate the extent to which exchange rate changes affect Dutch exports and imports. To this end, Marshallian demand functions are specified for exports and imports, under the assumption of imperfect substitution between domestic and foreign goods.

Exports

35. Exports are likely to depend on both economic activity in export markets as well as relative export prices:

Xd(Yw,Px,Px*)=Ywμ(Px/Px*)γ(1)

where Xd is the volume of exports demanded, Yw is the economic activity in export markets and Px/Px* is the relative price of exports in domestic currency. Taking logarithms and adding a constant term yields:

xd=α+μyw+γ(pxpx*)(2)

The profit function for exporters is:

Π(px,xd,ulc)=(pxulc)xd(yw,px,px*)(3)

where ulc denotes unit labor costs. Under the assumption that unit labor costs do not depend on the level of production, and that px*-ulc, maximization gives the following pricing rule:

pxpx*=12γ+12(ulculc*)(4)

This pricing rule can be substituted into the export demand function in equation (2) to obtain an alternative specification:

x=α*+μyw+ω(ulculc*)withω=γ2,α*=α+12(5)

These specifications for exports were estimated for total exports, exports of goods, and manufacturing exports (Table 9). For Yw export market volume was used, which is given by the trade weighted average of total imports, in volume terms, of all trading partners to which the Netherlands exports.

36. Using relative export prices, estimated price elasticities of exports range between 0.22 percent for total exports and some 0.57 percent for manufacturing exports.9 10 Activity elasticities are 1 percent for total exports and 1.25 percent for manufacturing exports.11 The lower elasticities for total exports might be due to the high share of agricultural products in total exports, which are usually characterized by low price-and income elasticities. The unitary elasticity obtained for total exports suggests that the market share of exports should be relatively stable over time, a trend confirmed by the data for total exports.

Table 9.

Netherlands: Export Volumes: Annual Data, 1970–96

article image
Sources: OECD, Analytical Database and Economic Outlook; and Fund staff calculations.Note: Estimation by Least Squares.XGSV: Volume of Exports, Goods and ServicesXGV: Volume of Exports, GoodsXMV: Volume of Exports, ManufacturingXGVMKT: Market Growth Index: Export, Manufacturing GoodsPXMDRA: Relative export prices in manufacturing, calculated using export weights.ULCMDRA: Relative unit labor costs in manufacturing, calculated using export weights.All variables are in logarithms.Standard errors are in parentheses.*: significant at 10 percent**: significant at 5 percent***: significant at 1 percent

37. When relative export prices are replaced by relative unit labor costs, the price elasticity estimates reduce by almost half. This suggests that, in accordance with equation (4), almost half of a change in relative labor costs is absorbed by profit margins.12 The estimations would suggest that a one percent decline in relative labor costs leads to an increase in total export volumes by 0.11 percent. A similar calculation for manufacturing exports implies that a 1 percent decline in unit labor costs leads to a 0.29 percent increase in export volumes.

Imports

38. For import demand the following functional form was adopted:

Md(Y,Pd,Pm)=Yυ(Pd/Pm)ρ(6)

where Md is volume of imports demanded, Y is the a scale variable denoting domestic activity, Pm is the price of imports and Pd is the domestic price of import-competing goods. In the empirical analysis alternative measures of relative prices were used, viz. the ratio of domestic demand deflator to import prices; the relative consumer price index, relative unit labor costs in manufacturing; and the real effective exchange rate based on consumer prices.

39. It was not possible to find an empirically meaningful relation between exchange rates and imports (Tables 10, 11, 12 and 13)13 This is likely the result of multicollinearity of the exogenous variables in the equation, which are both influenced by labor costs. A decline in labor costs results in a real exchange rate depreciation; it also results in a decline in the labor income share, and this shift in income from labor to capital reduces spending and hence imports. However, real effective exchange rate depreciation are likely to have a quite important impact on imports, both through their price effect (making imports more expensive) as through their income effect (shifting income from labor to capital).

Table 10.

Netherlands: Import Volumes: Annual Data, 1970–96

article image
Sources: OECD,. Analytical Database and Economic Outlook; and Fund staff calculations.Note: Estimation by Least Squares.MGSV: Volume of Imports, Goods and ServicesMGV: Volume of Imports, GoodsMMV: Volume of Imports, ManufacturingFDDV: Volume of Domestic DemandCPIDR: Relative Consumer Price IndexPDD: Ratio of Imports Prices (Goods and Services) to domestic demand deflator

All variables are in logarithms.

Standard errors are in parentheses.

* : significant at 10 percent

** : significant at 5 percent

*** : significant at 1 percent

Table 11.

Netherlands: Import Volumes: Annual Data, 1970–96

article image
Sources: OECD, Analytical Database and Economic Outlook; and Fund staff calculations.Note: Estimation by Least Squares.MGSV: Volume of Imports, Goods and ServicesMGV: Volume of Imports, GoodsMMV: Volume of Imports, ManufacturingFDDV: Volume of Domestic DemandEREER: Real effective exchange rate based on consumer prices.ULCMDRA: Relative unit labor costs in manufacturing, calculated using export weights.All variables are in logarithms.*: significant at 10 percent**: significant at 5 percent***: significant at 1 percent
Table 12.

Netherlands: Import Volumes: Differenced Data, 1971–96

article image
Sources: OECD≫ Analytical Database and Economic Outlook; and Fund staff calculations.Note: Estimation by Least Squares.D represents the first difference of the variable.Standard errors are in parentheses.*: significant at 10 percent**: significant at 5 percent***: significant at 1 percent
Table 13.

Netherlands: Import Volumes: Differenced Data, 1971–96

article image
Sources: OECD, Analytical Database and Economic Outlook; and Fund staff calculations.Note: Estimation by Least Squares,D represents the first difference of the variable.Standard errors are in parentheses.*: significant at 10 percent**: significant at 5 percent***: significant at 1 percent

Trade balance

40. A regression of exchange rate on the trade balance would suggest that a 1 percent increase in relative unit labor costs would reduce the trade balance by 0.33 percent for total goods and services, and by 0,31 percent for total goods (Table 14). With strong multicollinearity between the exchange rate and domestic demand, it was not possible to include demand variables; thus, this result should be interpreted with care.

Table 14.

Netherlands: Trade Balance: Annual Data, 1970–96

article image
Sources: OECD, Analytical Database and Economic Outlook; and Fund staff calculations.Note: Estimation by Least Squares.TBGSV: Volume of Trade Balance, Goods and ServicesTBGV: Volume of Trade Balance, GoodsTBMV: Volume of Trade Balance, ManufacturingULCMDRA: Relative unit labor costs in manufacturing, calculated using export weights.All variables are in logarithms.* : significant at 10 percent** : significant at 5 percent*** : significant at 1 percent

REFERENCES

  • Bakker, Bas B., Saving in the Netherlands, Sources, Trends and Mobility, Groningen (1993).

  • Central Planning Bureau, Central Economic Plan 1996, pp. 286310.

  • De Haan, L., Corporate Dividend Policy under Asymmetric Information: An Empirical Study for the Netherlands, De Nederlandsche Bank, MEB Series No. 3 (1994).

    • Search Google Scholar
    • Export Citation
  • Deppler, M.C. and D.M. Ripley, 1978, “The World Trade Model: Merchandise Trade,” Staff Papers, International Monetary Fund, Vol. 25 (MONTH), pp. 147206.

    • Search Google Scholar
    • Export Citation
  • Goldstein, M., and M.S. Khan, 1985, “Income and Price Effects in Foreign Trade,” in Handbook of International Economics, (eds by R.W. Jones and P. B. Kenen (Elsevier).

    • Search Google Scholar
    • Export Citation
  • Halikias, Ioannis, 1996, “Long-Term Trends in the Saving-Investment Balance and Persistent Current Account Surpluses in a Small Open Economy: The Case of the Netherlands,” IMF Working Paper 96/42 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • U.S. Department of Commerce (1995), Survey of Current Business (August 1995), pp. 53114.

1

Prepared by Bas Bakker.

2

Securities are usually valued at market prices; direct investment, on the other hand, is typically valued at historic prices.

3

The capital income share is the complement of the labor income share.

4

This is in accordance with previous analysis of the staff, which attributed the increase in the saving-investment balance to the slowdown of real wage growth.

5

Tax treatment of contractual saving is typically more favorable than tax treatment of other forms of saving. While pension benefits are taxed, pension premiums are deductible, and investment income of pension fund is untaxed. Since pensions are taxed at a lower rate than at which premiums are deducted, this entails a substantial fiscal subsidy. For savings channeled through banks, no such subsidy exists: premiums are not deductible, and interest income is taxed. Tax treatment of annuities and life insurances can be even more favorable than those of pensions, and fiscal constructions abound that combine borrowing of funds with the purchase of life insurance.

7

For empirical evidence that dividends of Dutch companies are sticky, see, for instance, Bakker (1993) and de Haan (1994).

8

Prepared by Meral Karasulu.

9

Goldstein and Khan (1985) report a selection of long-run price elasticities between 0.59 and 2.72 for total Dutch exports.

10

Statistical properties of the data enable the estimation of a stable relationship only for total exports and exports of goods when logarithmic levels of the data were used. The equation for manufacturing exports appears to represent a spurious relationship based on Augmented-Dickey Fuller test results. However, due to power problems of these tests, caution should be exercised in interpreting these results.

11

Deppler-Ripley, (1978) report a long-run activity elasticity of 0.65 for manufacturing goods. Goldstein and Khan (ibid) recount estimates ranging from 0.85 to 1.91 for total Dutch exports.

12

The estimated response of relative export prices to changes in cost competitiveness (Box 2) is slightly lower than ½, but the restriction of a 0.5 coefficient can not be rejected.

13

Regression results obtained from first differences of the data confirm that none of the relative price terms have economically meaningful coefficient estimates.

Kingdom of the Netherlands—Netherlands: Selected Issues
Author: International Monetary Fund