The Dutch Disease

Ronnie O'Toole - Senior Sophister

The recent good fortune of the Irish economy is a frequently discussed issue. Yet, behind this good fortune, could there be the signs of an incipient Dutch Disease? Ronnie O'Toole discusses the nature of the disease and its implications for applied economic analysis.

Introduction

In 1959 a large reservoir of natural gas was discovered in the Netherlands, which by 1976 earned that country revenues of some $2 billion in addition to an estimated $3.5 billion of savings in imports. By the mid 1970s, gross corporate investment had fallen by 15% since the start of the decade, while employment in manufacturing had declined by 16%. The total level of unemployment had risen from a modest 1.1% to 5.1%, while the share of profits in national income which had averaged 16.8% in the 1960s had fallen to 3.5% in the first half of the 1970s. While the first oil crisis had a devastating effect on most of the western industrial base, why did The Netherlands, with its new-found fortune in natural gas, fare worse than most?

This process of de-industrialisation of the existing manufacturing base was attributed to the upward pressure that the energy discovery placed on the Guilder and the wage rate, and was dubbed the Dutch Disease. Since then, the term's use has widened considerably to encompass any situation whereby a country's apparent good economic fortune ultimately proves to have a net detrimental effect.

Given the importance of this theory to the non-tradable sector, the salient features of Salter's (1959) non-tradable model will be outlined first. I will then present the Corden and Neary (1983) core model, showing how a boom in one sector can adversely affect the output of another. Following this, I will discuss two ways in which the Dutch Disease could have impacted on Ireland. I will conclude with a discussion on the relevance of the theory in practice.

The Basic Dutch Disease Model

The Salter Non-Tradable Model

Assume a small open economy that is a price taker in international markets and therefore cannot influence the world terms of trade. Define non-tradables as goods that face such transport costs as to prohibit trade, the classical example being a hair cut. Falvey (1994) identifies three consequences of the existence of non-tradables:

Before analysing the economics of the Dutch Disease, it is useful to first examine the effect of including non-tradables (N) in the analysis when there is a once-off outward shift of a country's (relative) capacity to produce tradables (T). Such an analysis can proceed by assuming that there is a shift in the world's terms of trade, with exportables (X) rising in price relative to importables (M). Figure 1a represents the appropriately modified version of Salter's model, and uses his convention of measuring T in units of M, with the terms of trade as the conversion rate between exportables and importables.

Initially the country is in both internal and external equilibrium (point e), with MRS = MRTS and with the foreign exchange rate (slope of rr) tangential to both the Production Possibilities Frontier (PPF) and the community indifference curve.

The change in the terms of trade results in an outward shift in the PPF as the country can produce more X in terms of M. This is the producer terms of trade effect. The community indifference curve also shifts upwards - more T (in terms of M) is needed to buy the same basket of goods. In other words, the price of M, in terms of M is obviously the same, but the cost of X is higher. This is the consumer terms of trade effect.

We can be sure that the net national terms of trade effect (the gap between the new PPF and IC) will be positive by recourse to Salter's analogy of a producer who consumes some of his own produce. A rise in the price of the good will prove beneficial to him as producer, while it will prove detrimental to him as consumer. Unless he consumes all of his own produce, the rise must be beneficial in aggregate.

As can be seen from Figure 1b, at the existing exchange rate the economy is producing ep while consuming ec and is thus running a balance of trade surplus while suffering over-employment at home. The position of ep is a case of Ryscynskis result that was published four years previously but which was not mentioned by Salter.

There are two steps necessary to restore simultaneous internal and external balance.

Hence the effect of the non-traded sector is to increase the price of exportables so as to decrease some of the new-found competitive advantage and also to decrease the price of importables, squeezing domestic import-substituting producers.

The Corden and Neary Core Model

An obvious improvement on Salter's model would be to give a more explicit explanation as to the importance of supply side factors in determining equilibrium. Corden and Neary (1983) did this, and were further able to explain what happens to factor shares in each of the sectors involved. The paper was written at a time of high output of North Sea oil, strong levels of sterling and stubbornly high wage demands in the UK. It is in effect the non-monetary companion piece of Corden (1981), which had been published some 18 months previously, and was written in an attempt to disentangle the impact of the Conservative government's monetary squeeze in the early 1980's, and any Dutch Disease effects of North Sea oil on the existing manufacturing base.

They based their core model on the assumption that there were three industries, namely a non-traded sector (N), a lagging sector (L) and a booming sector (B). Furthermore, they employed the factor specific model, assuming that labour is perfectly mobile between all three sectors but that the capital in each is sector specific.

The model can be explained with the aid of figure 2a and figure 2b, the first of which is constructed along the same lines as Salter, but taking the output of the lagging sector as numeraire. The second shows the labour demand schedules of N, L and T, where T is the total traded sector. Assume that the PPF curve shifts outwards, this time because of a once-off technological improvement in the booming sector. The authors identify two effects, the resource movement effect and the spending effect which are associated with the de-industrialisation of the traditional manufacturing sector.

Resource Movement Effect: This is the effect whereby the increased productivity in the booming sector pushes the equilibrium wage rate up, bidding the mobile factor out of the production of L.

At the given terms of trade, the boom raises production from the initial equilibrium of e to the point ep in both diagrams. In figure 2b, B's curve shifts upwards, thus also shifting upwards the aggregate tradables curve, raising the equilibrium wage rate from W0 to W1. The resulting reduction of output in L that is associated with the reduced use of labour is called direct de-industrialisation.

Now consider figure 2a below. At this stage we must make the further assumption that non-tradables have zero income elasticity, thus we can isolate the resource movement effect. While output has increased from e to ep, desired consumption of non-tradables has remained unchanged at b. To return the system to equilibrium, the opportunity cost of N in terms of L must rise, meaning the currency must appreciate. This will have the dual effect of increasing production while simultaneously reducing consumption, so the equilibrium point will lie somewhere to the right of ep.

Corden and Neary argue that this equilibrium must lie somewhere to the left of b because "the fall in services sector's output cannot be reversed". A less vague requirement would be simply to make the sufficient but not necessary assumption that non-tradables have non-positive price elasticity over the range in question.

In summary, the rise in the price of B has shifted its labour demand schedule upwards, bidding the mobile factor out of the production of both N and L. It should also be noted that the currency appreciation which returned the system to equilibrium somewhere between ep and b is an example of the indirect de-industrialisation which is more prominent in the spending effect.

Spending Effect: Some of the increased income created by the boom will be spent on non-tradables, which will push wages in that sector upwards. This also has the effect of bidding labour out of the production of L and B.

To isolate the spending effect, it is assumed that the energy sector does not use any labour, and hence the only effect of the boom is to push the PPF outward. Assuming that services are normal in aggregate, the point of desired consumption moves along some locus in figure 2a such as O to ec. At the initial exchange rate, there is an excess demand for N. Desired production is at b where the same amount of N is produced pre- and post- boom. The standard economic result of a rise in the price of N (an appreciation of the currency) now applies, shifting its labour demand schedule in figure 2b from Lno to Ln1, reducing the output of both of the tradable goods. This is the so-called indirect de-industrialisation of L.

The total impact on the output of N is ambiguous. If the spending effect dominates, then the equilibrium level of output will be somewhere to the right of b, while if the resource movement effect dominates, equilibrium will be to the left of b.

Effect on Factor Incomes

The resource movement effect results in a rise in the real wage in terms of traded goods. The fall in the output of non-tradable is associated with a rise in the wage in terms of non-tradables. Therefore the resource movement effect must unambiguously raise the real wage measured in terms of the full basket of commodities.

The spending effect, however, is ambiguous. The real appreciation of the currency as a result of increased spending on N results in a rise in the wage in terms of traded goods. However, the rise in the output of N is associated with a fall in the wage in terms of N, so the net impact of the spending effect depends on its relative magnitude and the resource movement effect.

Finally, the effect of the boom on the sector specific factors must be considered. The profitability of L must unambiguously fall, while profitability in N will rise on the spending effect and fall as a result of the resource movement effect.

The most interesting sector, however, is the booming sector. While profitability will rise because of the resource movement effect, it will fall as a result of the spending effect. This raises the possibility that the benefits of the boom will be so dispersed that the factor profitability of the booming sector might fall.

Ireland and the Dutch Disease

From this central model, a number of variants have appeared which isolate the Dutch Disease under a number of differing assumptions, two of which are particularly relevant to Ireland.

The Effect of Multinationals

In relation to the Irish case, Barry and Hannon (1995a) develop the Dutch Disease idea in response to a 'Leontieff' type paradox found in relation to Irish trade and Spanish trade. These two peripheral EU countries both had very high unemployment rates and were hence relatively labour abundant. Despite this, both countries had a revealed comparative advantage in the production of modern manufactures.

Two other European countries at a similar stage of economic development, namely Greece and Portugal, concentrated resources in labour intensive industries, and both had unemployment rates well below that of either Spain or Ireland. The implication was, of course, that the policy of attracting multinational companies had forced the equilibrium wage up and, à la Dutch Disease, the unemployment rate had increased.

They base their model on the assumption that the supply of the factors of production are close to being perfectly elastic. In the case of the labour market, any upward tendency in wages will result in an increase in the work force through immigration flows, reduced unemployment and changes in female participation rates. In the case of capital, Irish interest rates are very close to prevailing international rates when a small risk premium is allowed for.

Therefore in their model the resource movement effect becomes redundant, with the spending effect becoming the mechanism by which de-industrialisation occurs. The shock to the economy is a capital grant, with the assumption that all capital is ultimately repatriated.

Their result is that in a scenario of flexible exchange rates, the first order effect of the grant is to strengthen the currency, rather than cause an overall surplus in the balance of payments as would be the case with fixed exchange rates. This de-inflationary impact on the existing manufacturing base is compounded by the secondary effect of a fall in the price of non-tradables.

If the wage rate is not perfectly flexible, these deflationary shocks will result in a rise in unemployment which may be larger than the increase in employment due to the original inflow of multinationals. In the longer term, even when the wage rate has had time to adjust, the effect of hysteresis results in the initial fall in unemployment in the indigenous sector becoming permanent. Note that this latter idea is generalised below in the context of EU transfers.

EU Transfers

An alternative suggestion put forward by Krugman (1987) focuses on the interaction of the evolution of comparative advantage over time through 'learning by doing' and the Dutch Disease. While this theory will be sketched using a transfer from abroad as the relevant shock, it could as easily be explained using the more standard disturbance of a discovery of energy reserves.

The argument goes that the increase in productivity due to EU transfers to Ireland could have the disadvantageous effect of an upward pressure on wage rates which might be sufficiently great to overtake any productivity advance that the transfer might cause. This forces some sectors of the economy to relocate production elsewhere. The repercussions for the domestic economy is that the loss of 'learning by doing' due to the flight of particular sectors will result in declining domestic productivity as long as the transfer is maintained.

In the extreme case, those industries that left in the short run will not return, even when the transfer ends. The net result is that the country's market share and relative wage will decrease due to the temporary good fortune of EU transfers. That EU transfers could potentially have such an impact is clear given the huge scope of these payments, as shown in table 1 below.

This idea can be illustrated as in figure 3a, which shows the ratio of the home to overseas wage rate (W/w) on the Y-axis and the share of the world's tradables market supplied by domestic firms (p) on the X-axis. The trade balance equilibrium frontier (BB) is also shown. Now assume that a transfer T (in units of w) from the EU is received, which shifts the BB curve upwards by T(1-s)/s to B'B' in figure 3b.

From figure 3b it can be seen that as soon as the transfer is ended, B'B' returns to BB. However it is no longer economically justifiable to relocate production in the domestic economy, so the country is now faced with a reduced international market share in tradables of p1 instead of the original pre-transfer level of p0. This idea can be illustrated as in figure 3a, which shows the ratio of the home to overseas wage rate (W/w) on the Y-axis and the share of the world's tradables market supplied by domestic firms (p) on the X-axis. The trade balance equilibrium frontier (BB) is also shown. Now assume that a transfer T (in units of w) from the EU is received, which shifts the BB curve upwards by T(1-s)/s to B'B' in figure 3b.

The AA curve traces out the home country's share of the total tradable sector for each given (relative wage) level. It is downward sloping in that the higher the relative wage level, the lower will be the number of sectors in which the home country has a comparative advantage. Through external economies of scale and international specialisation, the AA curve might develop another step as in figure 3b.

From figure 3b it can be seen that as soon as the transfer is ended, B'B' returns to BB. However it is no longer economically justifiable to relocate production in the domestic economy, so the country is now faced with a reduced international market share in tradables of p1 instead of the original pre-transfer level of p0.

Table 1: Irish Receipts from the European Union

IRm Current Prices

1985

1990

1992

1994

1996*

FEOGA Guidance

56

94

147

131

152

ESF

141

129

277

277

291

ERDF

76

225

445

176

365

Other

19

7

11

84

162

FEOGA Guarantee

837

1287

1114

1174

1100

GROSS RECEIPTS

1129

1741

1994

1841

2069

Less contributions

213

283

353

506

633

Net Receipts

915

1458

1641

1336

1436

GNP

15262

22604

26452

30614

36545

Net receipts (%GNP)

5.9

6.4

6.2

4.4

3.9

*Estimates

Sources: (a) Central Bank of Ireland Annual reports, (1991-1996)

(b) Ireland's receipts by way of grants and subsidies from EU budgets, (1973-1996). European Commission, Brussels, (1996).

Conclusion

So far, a core model of the Dutch Disease and two theoretical ways in which Ireland could have contracted it have been given. Why should it matter, however, by which means de-industrialisation occurs? As comparative advantage evolves, there are always going to be winners and losers, and the end result of the economy's altered specialisation is surely all that is important. As Krugman (1987) points out, the only real worry is that the economy's good fortune, whether it is through an oil discovery, a transfer or a sharp change in the terms of trade, proves to be a temporary phenomenon.

This point aside, the de-industrialisation attributed to the Dutch Disease in the examples mentioned thus far cannot pass without comment. In the case of the discovery of oil in Holland, Corden (1984) blames the sluggish performance of the Dutch economy in the 1970s on the rise of the importance of the public sector which followed the gas discovery, resulting in a crowding out of investment by the private sector.

Finally, to believe Krugman's theory of the loss of 'learning by doing', in the case of Ireland, requires that any increase in real wage due to transfers (remembering the high supply elasticity of the Irish labour market) will outweigh all infrastructural, educational, and research and development benefits which accrued as a result of EU expenditures.

All in all, given the absence of the usual symptoms and given that it is not clear how we could have contracted it, Ireland is probably not suffering from the disease. Perhaps the Third World provides more fertile ground for Dutch Disease theorists. Many LDCs are typified by a large dependence on a very limited range of primary products. These products' prices in turn can be highly volatile, resulting in very large trade shocks. Such possibilities have given rise to numerous papers focussing on issues such as potential compensation measures in LDCs as a practical response to large shifts in prices.

 

Bibliography

Barry, F. and Hannon, A. (October 1995a) "Multinationals and Indigenous Employment: An 'Irish Disease'?". Economic and Social Review, vol. 27, (1), p. 21-32.

Barry, F. and Hannon, A. (1995b) "Multinationals and Indigenous Employment: An 'Irish Disease'?" in Working Paper, No. 95/13. University College, Dublin.

Collier C. and Gunning J.W. (1994) "Trade and Development: Protection, Shocks and Liberalisation" in Surveys in International Trade. Greenaway and Winters (eds), Blackwell: Oxford.

Corden W.M. and Neary, J.P. (1983) "Booming Sector and De-industrialisation in a Small Open Economy" in The Economic Journal, vol. 92, p. 825-848. Reprinted in W.M. Corden, "The Exchange Rate, Monetary Policy and North Sea Oil: The Economic Theory of the Squeeze on Tradables", Oxford Economic Papers, vol. 33, Supplement, p. 23-46. Reprinted in W.M. Corden, "International Trade Theory and Policy", Edward Elgar Publishing, 1992.

Corden W.M. (1984) "Booming Sector and Dutch Disease Economics: Survey and Consolidation" Oxford Economic Papers, vol. 36, p. 359-380. Reprinted in W.M. Corden, International Trade Theory and Policy. Edward Elgar Publishing, 1992.

The Economist (26 November 1977) Trade Theory and Policy.

Falvey (1994) "The Theory of International Trade" in Surveys in International Trade. Greenaway and Winters (eds), Blackwell.

Krugman, P. (1987) "The Narrow Moving Band, the Dutch Disease and The Competitive Consequences of Mrs. Thatcher on Trade in the Presence of Dynamic Scale Economies" in The Journal of Development Economics. vol. 27, p. 41-55.

Salter, W.E.G. (1959) "Internal and External Balance: The role of price and expenditure effects" in The Economic Record. vol. 35, p. 226-238