The GCC's Fixed Exchange Rate:
October 2009
The GCC’s Fixed Exchange Rate: A Major Anomaly for OCA Analysis
Thomas D. Willett* Khalfan M. Al-Barwani** Sherine M. El Hag*** School of Politics and Economics Claremont Graduate University California
*Corresponding Author Director, Claremont Institute of Economic Policy Studies and Horton Professor of Economics Claremont Colleges [email protected] **Part time lecturer at Sultan Qaboos University *** Lecturer California State University, Fullerton [email protected]
Abstract: From the stand point of traditional OCA analysis, the GCC countries are a major anomaly. They successfully maintained pegged exchange rates among themselves for two decades while failing to meet a number of major OCA criteria. Contrary to the optimistic predictions of some of the most enthusiastic advocates of endogenous OCA analysis, intra-regional trade has remained low and there is little synchronization of their business cycles or harmonization of their fiscal positions. We argue that their success in operating a fixed rate regime has been due to a combination of high oil revenue, the heavy use of foreign workers, and the limited extent to which their economies rely on market forces. These are conditions that are not met by most other countries.
1. Introduction The Gulf Cooperation Council (the GCC) countries have pledged to form a monetary and currency union1. Toward this end, the six GCC countries officially pegged their respective currencies both in terms of de jure and de facto to the US dollar in 2003. By 1986, all the GCC member states, except for Kuwait 2, already had de facto pegs to the US dollar. The original date set for the union was 2010, but this deadline was recently extended. The official explanation for the delay was that more time was needed to complete the development of a common monetary and regulatory framework. There had been previous signs that the 2010 goal was unlikely to be met, with Oman announcing in 2006 a need for it to delay. However, there is still strong support for eventual monetary union among many of the GCC‟s national leaders and it remains the likeliest candidate for the creation of the next sizeable currency union. One of the most surprising characteristics of the GCC economies is that from 1981 until 2007 when Kuwait revalued its dinar and switched from a dollar to a basket peg, there were no major changes in intra group exchange rates, and the Kuwait revaluation was itself quite modest. Furthermore, despite the absence of capital controls, there were few instances of large speculative pressures until the current global crisis. 1
The GCC countries consist of Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates (UAE). 2
Until 2003, the Kuwaiti exchange rate regimes differed slightly from the rest of the GCC countries since the Kuwaiti Dinar was pegged to a basket of currencies of its main trading and financial partners. While the composition of the basket was not disclosed, the US dollar was believed to be the dominant weight since minor fluctuations occurred visà-vis the US dollar and subsequently the fluctuations vis-à-vis other GCC currencies was also fairly small (Sturm and Siegfried, 2005). 1
The reason for the intra group exchange rate stability likely resulted primarily from the GCC countries‟ common desires to peg to the dollar because of the pricing of oil in dollars3. What is surprising is that it was so successful despite the GCC economies failing to meet a number of the major criteria for being an optimal currency area (OCA). The answer to this puzzle we suggest lies in several unusual characteristics of the GCC economies that differ substantially from the standard assumptions implicitly underlying traditional OCA analysis. In section 2, we provide historical and institutional background on the GCC. In section 3, we briefly review how the GCC economies fit with the major OCA criteria. We find that in general they fail to meet most of these criteria. For example, while they all are highly open economies, they have relatively little trade with each other and this is what is relevant for the economic costs and benefits of forming a currency area. In section 4, we provide a suggested explanation for why the GCC‟s pegged rate regime worked so well for so long despite its failure to meet the traditional OCA criteria. In sections 5 and 6, we consider other possible explanations. In section 5, we critique Willem Buiter‟s recent critique of OCA analysis and in section 6, we consider the possibility suggested by the recently developed endogenous OCA analysis that the fixing of GCC exchange rate unleashed powerful forces to move the GCC countries much closer to meeting the OCA criteria. We do not find evidence of strong endogenous effects on trade flows, business degree synchronization, or fiscal policy coordination. Section 7 offers concluding comments. We suggest that the success of exchange rate fixity within the GCC provides
3
We are indebted to a referee for suggesting this point. 2
little promise that the macroeconomic costs of fixed exchange rate would be low for countries that do not share these characteristics.
2. Historical and Institutional Background In May 1981, the Arab Gulf region countries: Bahrain, Kuwait, Oman, Qatar, Saudi Arabia and the United Arab Emirates, formed the Cooperation Council for the Arab States of the Gulf (GCC countries). The main objectives of the GCC include strengthening the economic cooperation among Member Countries, harmonizing their economic and financial policies, promoting the private sector, and achieving a common market and monetary union. There are a number of common characteristics among the GCC counties. These include heavy reliance on hydro carbon resources for their domestic and external economies, geographical proximity, small private sectors and dominant public sectors, extensive reliance on foreign labor force in the private sector, common language and culture, and small populations except for Saudi Arabia where about 70% of the estimated 35 million combined population of the GCC reside. The GCC charter, known as “Charter of the Cooperation Council”, established the main bodies of the GCC, which consist of the Supreme Council, the Ministerial Council and the Secretariat General. The Charter also laid down the compositions of these bodies and their functions. They include the intergovernmental decision-making and consultative bodies with the Supreme Council and the Ministerial Council at its core, the Secretariat General as the main supranational institution, plus a number of specialized regional agencies.
3
The Supreme Council is the highest authority of the governmental bodies where all main decisions are made. This Council consists of the Heads of State of the six GCC member states and its presidency is rotated on annual basis. The Supreme Council, which meets twice a year, has the task of providing policy direction, reviewing the reports and recommendations submitted by other bodies and appointing the Secretary General of the GCC. Foreign Ministers or other ministers of the member states form the Ministerial Council. This Council is tasked with policy proposals, preparing recommendations, and conducting studies in all concerned areas. A number of specialized committees also work under the Ministerial Council. The most important committee with regard to economic integration is the Committee for Financial and Economic Cooperation, which includes the Ministers of Finance and Economies. The various ministerial committees direct a number of sub-committees to conduct studies, write proposals and coordinate national policies at a technical level. Resolutions passed by the Supreme Council as well as the Ministerial Council are on the basis of unanimity for major issues and by majority rule for procedural matters. Cooperation among the GCC central banks takes place under the Monetary Agencies and Central Bank Governor‟s Committee. This committee meets twice a year and reports to the Committee for Financial and Economic Cooperation and its subcommittees, which consist of the Banking Supervision Committee, the Payment Systems Committee, the Monetary Union Committee, and the Training Committee.
4
According to the Charter, the Secretariat General works independently and for the common interests of the member states, and unlike the two other regional bodies, it does not consist of representatives of the member states. Its headquarters is located in Riyadh, Saudi Arabia and the Secretary General, who heads this institution, serves a three-yearterm subject to one renewal. The functions of the Secretariat General include administrative preparation, follow-up of the meetings of the intergovernmental bodies, preparation of studies and reports on matters pertaining to the cooperation objectives, follow up the implementations of resolutions and recommendations, and draft resolution on common legislation. The Secretariat also acts as a representative of the six member states in certain international forums. Additionally, this body has been assigned the task of settling disputes regarding the implementation of the GCC Economic Agreement and its ratified decisions, including claims raised by private parities. In the event the Secretariat General fails to settle a dispute, it can refer the matters to the GCC Commercial Arbitration Center, a judicial body. The six member states have also set up some specialized agencies to deal with issues related to patent rights, commercial disputes, and product standardization. In November 1981, the GCC established an economic agreement to set the stage for full economic integration (Laabas and Liman, 2002). In 1982, the goal to create a unified currency was declared by the six states. In 1983, a free trade zone was formed as a first step toward their economic integration. However, it took the GCC states sixteen years to reach a custom union agreement in 1999. In 2000, the GCC countries agreed to adopt a common exchange rate peg as a step toward the realization of a common currency in 2010. In 2001, an accord on a joint custom tariff of five percent was reached
5
and implemented two years later. In 2002, the US dollar was selected as the intermediate peg for the six member states‟ currencies before the proposed unified currency comes into effect. Officially, the six GCC member states began to peg their respective currencies to the US dollar in 2003 as a step towards adopting a common currency projected to take place in 2010. Unofficially, this policy was adopted much earlier. In fact, by 1986, all of the GCC member states, except for Kuwait 4 had initiated their de facto pegs to the US dollar. In early 2004, the GCC heads of central banks agreed in principle on key convergence criteria for entry into the common currency. These included the size of budget deficits, inflation rates, interest rates, foreign reserves and the ratio of public debt to gross domestic product. There is not general agreement on the primary reasons for the GCC‟s pursuit of a monetary union. While some have suggested economic explanations based on OCA considerations ( see, for example, Al-Jasser and Al-Hamdy(2003), others have suggested that both the economic costs and benefits would be modest, see, for example, Sturm and Siegfried (2005) and still others have concluded that important OCA conditions are not met (Abu-Qarn and Abu-Bader (2008). In general, we find that while economic arguments are often used, the actual formation of currency areas is usually based more on geopolitical considerations.
4
Until 2003, the Kuwaiti exchange rate regimes differed slightly from the rest of the GCC countries in the sense that the Kuwaiti Dinar was pegged to a basket of currencies of its main trading and financial partners. While the composition of the basket was not disclosed, the US dollar was known to be the dominant weight since only minor fluctuations occurred vis-à-vis the US dollar and subsequently the fluctuations vis-à-vis other GCC currencies were also fairly small (Sturm and Siegfried, 2005). 6
This was the case in Europe with the creation of the Euro and was likely important also with the GCC5. Also likely is the dynamic observed in Europe of pushing from one stage of economic integration to another. While the experience of the North America Free Trade agreement among Canada, Mexico, and the United States shows that countries can take major steps to increase trade integration without contemplating monetary integration, in Europe there were strong political pressures to continue on to monetary union. Since there is relatively little trade among GCC countries, it seems likely that there was much less pressure for monetary integration from economic interest groups then there was in Europe. Since integration efforts were begun the GCC countries made considerable progress on reducing inflation rates and this was maintained until quite recently. This resulted in considerable convergence of inflation rates. Rates of economic growth have shown considerable divergence, however. For example, over the period 1975 to 2003, average growth rates range from at high of almost 7 percent in Oman to a low of below 1 percent in Kuwait. The largest country, Saudi Arabia, averaged 2.4 percent (Abu-Qarn and AbuBader,2008).
3. The GCC Anomaly and a Suggested Explanation The theory of optimum currency areas lays out the factors that explain the costs and benefits of countries adopting hard fixed (i.e. joining a currency area) versus flexible rates. The major cost of joining a currency area is giving up independent monetary and exchange rate policies. This cost is low if the domestic economy is quite flexible and 5
On Europe see the analysis and references in Willett(2003 ). On the role of geopolitical considerations within the GCC see Popescu and Mustafa(2001). 7
could be high if the economy is rigid. Furthermore, if the economy is highly open, exchange rate adjustments lose their effectiveness, while the importance of the external sector grows relative to the internal one. Thus, large closed economies with domestic rigidities are ideal candidates for flexible rates while small open economies with flexible domestic labor and product markets are ideal candidate for fixed rates. This assumes, however, that the country in question has a stable partner or group of countries to which to fix its exchange rate. Thus, patterns of shocks and correlation of macro economic conditions could be important as well. 6 In recent years, a major research industry has developed applying OCA criteria to various countries or group of countries. Of necessity, most of these empirical studies have looked at attributes of countries under various forms of flexible rates and tried to predict how well the fixing of exchange rates would work.
The case of the GCC
countries, however, provides a natural experiment of how well they operated under fixed rates. While not all the major OCA criteria point in the same direction (for a summary of previous studies of the GCC see Abu-Qarn and Abu-Bader(2008), our reading is that the most relevant OCA conditions are far from met. As Abu-Qarn and Abu-Bader conclude, "In line with most of the previous studies, the three methods employed here provide no evidence of the readiness of the GCC members to establish a lasting and well functioning currency union." (2008,p 629). What is surprising is that despite failing to meet many of these criteria, the GCC countries have successfully maintained a regime of fixed exchange rates for two decades without the widespread use of controls. How can this be?
6
On the development of OCA theory pioneered by Mundell, Mckinnon, and Kenen and surveys of recent developments see Tavlas (1993) and Willett (2003). 8
At first glance, all of the GCC countries except Saudi Arabia meet the major OCA criteria of being small open economies. All the countries except Saudi Arabia have ratios of imports to GDP of greater than 30%. What is relevant for currency formation, however, is how open the prospective members are to trade with each other and on this score the GCC countries score quite low. Only Oman had more than 10% of its exports go to other GCC countries. In 2003, the average ratio of intra group to total exports was only 5 per cent. Low intra- area trade was the reason why the three small Baltic States (Estonia, Latvia, and Lithuania) did not seriously consider forming a common currency area after the break up of the former Soviet Union. While there is not enough data available to do formal analysis, it is generally judged that levels of factor mobility by GCC citizens are low, as is the degree of wage flexibility7. However, GCC labor markets are “highly unusual” to use William Buiter‟s (2007) phrase. They have more flexibility than these conditions suggest because of the heavy use of expatriate workers. These expatriates work primarily in the private sector, while nationals work primarily in the public sector which is characterized by wage rigidity, immobility, and inflexibility. Foreign labor constitutes about 55% of total employment in Bahrain and more than 80% in the UAE, Kuwait and Qatar 8. This combination of high levels of protected jobs for the locals and large movements of foreign workers cushion the domestic effects of macroeconomic fluctuations despite the domestic rigidities. While this process has provided considerable adjustment in the past, Willem Buiter (2007) questions whether it will continue to operate 7
For more detailed analysis of the GCC economies, see Jadresic, 2002, Laabas and Liman (2002), and Sturn and Siegfried (2005) 8 The expatriate labor is usually from India, Philippines and Indonesia, other MiddleEastern nations, Europe and USA. 9
as well in the future. The labor mobility among the expatriates “can be a substitute for the removal of legal and administrative obstacles to cross-border labor mobility by resident workers, native and expatriate, which would [be] the standard way to use labor mobility as a substitute for exchange rate flexibility. It is unclear how long the current migration paradigm of the GCC countries can survive. At some point the „guest workers‟ will demand greater rights and will no longer be willing to accept passively to act as the residual in the GCC labor markets” ( p27) .
4. The Anomaly Explained Traditional OCA analysis was developed within the context of economies with large private sectors and only intra- area labor movements. In such economies, substantial labor mobility and/or wage and price flexibility is needed to cushion the efforts of macroeconomic fluctuations under fixed exchange rates since the use of independent monetary policies is taken away. The GCC countries have been able to compensate for the lack of such domestic flexibility to a considerable degree through the mobility of expatriates described above. Combined with limited need for adjustment by nationals in the private sector, high levels of revenue from oil sales( see table 1) have given governments the capacity to effectively protect the jobs of domestic workers without severe fiscal strains.
10
Table 1 Oil Dependency of the GCC and its Member Countries S. ARABIA
UAE
GCC
56.80
34.90
28.10
36.00
78.40
64.20
80.60
75.30
79.30
64.50
34.50
65.50
38.80
66.70
BAHRAIN
KUWAIT
OMAN
QATAR
GDP*
25.70
45.90
43.10
Govt. Revenue**
73.00
91.50
Exports***
66.70
83.80
Oil as Share of
Sources: European Central Bank (ECB), Arab Monetary Fund (AMF), GCC central banks, Internationa Mnetary Fund (IMF), and Institute of International Finance (IIF) * Oil and gas sector's share of GDP as a % in 2001 ** Oil revenue/total government revenue as a % (include gas revenue for Bahrain) in 2000 *** Oil and oil products' share of total exports as a % in 2004
This insulation of nationals from the effects of macroeconomic fluctuations has allowed the GCC countries to achieve considerable harmonization in their rates of inflation until recently (see chart 1 below). This was done without generating the types of strong domestic unemployment pressures that would be generated by giving up countercyclical monetary policies in the types of Keynesian economies analyzed in the initial OCA contributions.
11
Chart 1: GCC Inflation Rates 1980 – 2004
Data Source: WDI Note that Qatar data were unavailable for the period under review.
Of course even with harmonized inflation rates, the GCC countries‟ high levels of export instability generated by their heavy reliance on petroleum exports could generate pressures on intra group exchange rates. With fairly similar export structures, however, we would expect the symmetry of the major shocks to their export revenues to serve to reduce the tensions generated on intra group exchange rates. Surprisingly, however, AbuQarn and Abu-Bader(2008) find that the supply shocks to the GCC countries are asymmetric, while the demand shocks are more symmetric. They suggest the possible explanation that for the oil producers shifts in the international price of oil act as both supply and demand shocks. 12
This doesn‟t help with the GCC countries‟ abilities to maintain their external peg to the dollar, however. Here their high levels of oil revenues have allowed them to avoid much of the domestic macroeconomic adjustment that would have been needed in the absence of substantial scope for financing deficits. Again, standard OCA analysis assumes that countries must adjust to external imbalances that are not fairly quickly mean reverting. The GCC countries oil wealth substantially relaxed this constraint.
5. A Critique of Buiter’s Critique of OCA Analysis Our argument for the limited relevance of traditional OCA analysis for the GCC countries is quite different from the analysis by Willem Buiter (2007)9. He argues the much stronger proposition that “the theory of optimal currency areas…is almost entirely useless as a guide to the choice of currency regime in modern economies” (p15). With this generalization, we strongly disagree. Buiter argues that there are two key failures of the OCA approach. One is what he calls “chronic confusion between transitory nominal rigidities and permanent real rigidities” (p). We share Buiter‟s view that there is no favorable long run inflationunemployment trade-off such as was assumed in early Keynesian analysis ,but this needn‟t imply that nominal rigidities are not of sufficient importance to discretionary monetary policy to reduce the costs of adjusting to some types of economic shocks. Buiter‟s second argument is that “conventional OCA literature was designed for a world without endogenous capital mobility” (p16). His argument that capital flows can in some circumstances amplify disturbances is well taken. 9
While Buiter favors a GCC monetary union on economic grounds, he is skeptical that necessary political conditions are met. 13
However, there are also circumstances where capital mobility can dampen the effects of shocks. That high capital mobility can increase or reduce macroeconomic stability depending on both nature of the shock and the exchange rate regime is a basic result of Mundell-Fleming analysis (see Tower and Willett 1976). Buiter‟s real point is that destabilizing speculation can make the foreign exchange market an exogenous source of disturbance. While an unusual view for a new classic macro economist to hold, destabilizing speculation is certainly possible. This raises an important issue, but doesn‟t call for a full scale rejection of OCA analysis. One final comment on Buiter‟s analysis is in order. He describes all of the GCC countries as being small, open economies including Saudi Arabia. He uses a different criterion for openness than does conventional OCA analysis. He focuses on the ability of countries to influence their terms of trade. This is not necessary for exchange rate adjustment to be effective, however. The type of openness relevant for this is that there be enough non-traded goods and services in the economy so that a change in the nominal exchange rate can lead to a meaningful change in the relative price of traded and nontraded goods.
6. Endogenous OCA Analysis: Did the establishment of pegged exchange rates lead to a greater convergence in business cycle conditions? Another possible explanation for the success of the GCC fixed exchange rate regime comes from endogenous OCA theory that emphasizes that the act of fixing exchange rates may change the behavior of OCA criteria in ways that make fixed exchange rates more workable.
14
Fixing exchange rates should increase intra-regional trade and likely also increase the similarities of macroeconomic fluctuations, thus reducing the divergences among optimal national monetary policies. The creation of the euro area has been consistent with both of these predictions, although the concomitant increases in trade and business cycle correlations with other Europeans countries make it difficult to judge how much of these changes have been due to the creation of the euro per se10. For the GCC countries, however, the increase in intra regional trade over the fixed rate period has been modest. Table 2 describes the evolution of intra GCC trade.
10
See Willett, Permpoon, and Wihlborg (2007). Of course, the creation of a common currency would be expected to have stronger endogenous effects than the GCC‟s pegged rate regime. 15
Table 2: Intra GCC Exports & Imports as Shares of Total Exports and Imports from 1986 to 2003 FIVE-YEAR AVERAGES
19841988
19891993
1994-1998
1999-2003
GCC EXP % TOTAL EXP
GCC EXP % TOTAL EXP
GCC EXP % TOTAL EXP
GCC EXP % TOTAL EXP
15.621 2.584 5.252 5.480 5.891 4.526
6.640 1.997 8.180 6.536 6.751 5.275
8.837 1.613 12.543 5.400 7.437 7.066
7.002 1.525 11.984 4.937 5.195 6.457
1984-1988
1989-1993
1994-1998
1999-2003
GCC IMP % TOTAL IMP
GCC IMP % TOTAL IMP
GCC IMP % TOTAL IMP
GCC IMP % TOTAL IMP
44.169 4.234 22.567 9.983 1.881 9.371
41.631 4.443 26.988 11.129 1.907 10.066
37.244 9.638 27.429 13.295 3.050 9.440
25.610 11.629 32.205 14.970 2.951 8.340
COUNTRY BAHRAIN KUWAIT OMAN QATAR SAUDI ARAB UAE FIVE-YEAR AVERAGES
COUNTRY BAHRAIN KUWAIT OMAN QATAR SAUDI ARABIA UAE
Data Source: IMF Direction of Trade Statistics Intra-GCC Export as % of Total Export indexes are computed by the authors
To investigate the degree of synchronization of macroeconomic fluctuations, we computed the correlations among the first differences in the logs of real GDP over one, two, and three year time intervals. The recent literature has generally assumed that high correlations increase the case for a currency union since this would give rise to smaller differences in optimal discretionary monetary policies. Thus the constraint of adopting a
16
common policy would be less costly. We are doubtful, however, that monetary policy can normally be fine-tuned sufficiently to respond effectively to one year fluctuations. The combined lags in recognition and effects of changes in monetary policy on the economy are likely to be too long. So despite the common tendency in the literature to focus only on one year correlations, we believe that correlations over longer time periods of two to three years are more relevant for analyzing the scope of effective discretionary monetary policy responses. Indeed, it can be argued that over horizons of a year or less, low correlations would be superior as these would fit the automatic stabilization criteria emphasized in some of the OCA literature11. Table 3 shows the correlation matrices for 1970 through 2004.
11
See Tower and Willett ( 1976) 17
Table 3 GCC Growth Rate Correlation Matrices Table 3a: GCC One-year Growth Rate Correlations BAHRAIN
KUWAIT
OMAN
QATAR
SAUDI ARABIA
UAE
BAHRAIN
1.000
KUWAIT
0.022
1.000
OMAN
0.192
-0.130
1.000
QATAR
-0.255
0.248
-0.357
SAUDI ARABIA
0.438
-0.166
-0.159
0.244
1.000
UAE
0.623
-0.202
-0.057
-0.149
0.692
1.000
QATAR
SAUDI ARABIA
UAE
1.000
Table 3b: GCC Two-year Growth Rate Correlations BAHRAIN
KUWAIT
OMAN
BAHRAIN
1.000
KUWAIT
-0.012
1.000
OMAN
0.122
-0.213
1.000
QATAR
-0.445
0.187
-0.492
SAUDI ARABIA
0.496
-0.206
-0.247
0.226
1.000
UAE
0.716
-0.315
-0.042
-0.168
0.791
1.000
QATAR
SAUDI ARABIA
UAE
1.000
Table 3C: GCC Three-year Growth Rate Correlations BAHRAIN
KUWAIT
OMAN
BAHRAIN
1.000
KUWAIT
-0.062
1.000
OMAN
0.178
-0.288
1.000
QATAR
-0.384
0.131
-0.645
SAUDI ARABIA
0.625
-0.236
-0.267
0.148
1.000
UAE
0.741
-0.357
0.009
-0.126
0.869
1.000
1.000
Data source: IMF/IFS from 1970 to 2004 Computed by the Authors
The most striking feature of these calculations is the overall low magnitudes of most of the correlation coefficients. Moreover, a number of the pairwise correlation coefficients have negative signs. The highest correlation coefficients are between Saudi Arabia and the UAE and this is consistent in all three matrices. The other high correlation coefficients observed are those between Bahrain and the UAE with the highest magnitude of .74 for three-year growth rates and lowest of .62 for one year growth rates and between Bahrain and Saudi Arabia with the highest magnitude of 0.625 in the three-year growth
18
rate matrix and the lowest under the one-year growth rate matrix at 0.4 12. Across all three matrices there are consistent negative correlation coefficients between Oman and Qatar, Bahrain and Qatar, and of smaller magnitude, between Oman and Saudi Arabia. The GCC countries‟ average correlation coefficients for one-year growth rates are slightly higher than for the two and three-year averages. This conflicts with the ideal pattern which would display low or negative short run correlations for automatic stabilization and high long term correlations so that there would be less difference in optimal discretionary monetary policies response across countries. Too much should not be made of these correlation tables, however, since the correlations can vary substantially over time and not always along a steady path. This is illustrated in chart 2 which displays ten year rolling window calculations of the three year growth correlations of the five smaller GCC countries against Saudi Arabia. These give the ten year moving averages of these correlations. If there is any clear pattern, it is for both the high positive and high negative correlations displayed in the 1980s to converge toward lower correlations. In a number of cases, the correlations change signs in the latter part of the period. The drop in positive correlations with Saudi Arabia is especially strong for Oman and the UAE. Possible explanations include differences in the extent and speed of fiscal response to shocks from one country to another, and in the pace of economic diversifications.
12
The Bahraini economy has for years depended on Saudi grants of crude oil that it refines, consumes domestically and exports. 19
Chart 2: Three-Year GCC RGDP Growth, Ten-Year Rolling Window Correlations Three-Year Growth Rate: Ten-Year Busines Cycle Rolling Window Correlation, Between Saudi Arabia and Other GCC Countries 1970-2004 Correlations
1.20 1.00 0.80 0.60 0.40 0.20 0.00 -0.20 -0.40 -0.60 -0.80
BAHRAIN
KUWAIT
OMAN
QATAR
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
-1.00
UAE
Data source: IMF/IFS from 1970 to 2004
Given the similarity of major shocks through the oil markets, the generally low levels of business cycle correlations among the GCC countries is surprising. A part of the explanation is that while monetary policies have led to substantial convergence of inflation rates, fiscal balances have varied widely among the countries. Table 4 presents data on fiscal imbalances as a percent of GDP and changes in the ratio of government expenditure to revenue for the six GCC countries over various time periods and business cycles. Sturm and Siegfried (2005) find that while the underlying components of the budget deficits- government revenue and expenditures- exhibit high degrees of comovements, changes in the ratio of expenditure to revenue do not follow a similar pattern. Using a co-integration and error- correction modeling framework, Fasano and Wang (2001) find that government spending in the GCC countries is driven by revenue from oil, suggesting a pro-cyclical expenditure policy. However, the timing and extent of spending changes differs among countries and this results in different patterns shown in
20
table 4. Thus fiscal policy coordination cannot be used to explain the success in maintaining the GCC countries‟ internal pegs. Table 4: GCC Fiscal Policy Indicators Average form to
1975 1980
1980 1986
1986 1990
1990 1994
1994 1996
1996 1998
1998 2000
2000 2002
Deficit (-) or Surplus as % of GDP
-0.30
0.71
-4.92
-3.86
-3.65
-4.22
-2.73
n.a.
Change in expenditure/revenue*
0.90
-7.60
1.30
-1.50
0.40
-0.30
0.20
n.a.
Deficit (-) or Surplus as % of GDP
41.48
5.99
n.a.
n.a.
-3.40
1.77
4.45
19.34
Change in expenditure/revenue*
0.70
-0.90
-15.80
-2.00
0.20
-0.70
0.00
n.a.
Deficit (-) or Surplus as % of GDP
-1.59
-8.22
-9.24
-7.98
-7.67
-3.95
-6.52
n.a.
Change in expenditure/revenue*
0.50
7.20
0.00
-1.70
0.00
0.10
0.20
n.a.
Bahrain
Kuwait
Oman
Qatar Deficit (-) or Surplus as % of GDP
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
Change in expenditure/revenue*
0.60
0.10
0.10
-0.90
0.80
-0.20
0.10
n.a.
Deficit (-) or Surplus as % of GDP
n.a.
n.a.
n.a.
n.a.
n.a.
-3.06
-4.85
n.a.
Change in expenditure/revenue*
0.90
0.50
0.50
1.20
0.40
0.20
0.40
n.a.
Deficit (-) or Surplus as % of GDP
0.64
n.a.
n.a.
0.10
-0.14
0.19
n.a.
n.a.
Change in expenditure/revenue*
1.00
0.20
0.50
0.20
2.20
1.50
0.30
n.a.
234.80
-65.00
62.20
-30.60
27.70
-35.80
116.00
n.a.
Saudi Arabia
UAE
Change in average crude oil prices** Data source: IMF/IFS. Indices computed by the author *Indices provided by Fasano and Wang (2001)
** By Fasano and Wang (2001) based on average nominal price of UK Brent, Dubai, and West Texas Intermediate
Another factor often posited to help the workings of a currency area is financial integration which in turn would be encouraged by the establishment of fixed exchange rates. This doesn‟t appear to have been a major factor within the GCC countries. Despite the removal of most restriction on capital flows, there has been relatively little movement of financial capital within the GCC. The vast majority of capital flows are with outside countries and a substantial intra-GCC banking market is yet to appear (see Sturn and
21
Siegfried, 2005). Factors explaining the low movement of capital among the GCC countries include bureaucratic barriers and limited investment opportunities. The GCC financial markets are generally still in their infancy. In summary, there is little support for the hypothesis that strong endogenous responses to pegged exchange rates have been a major factor in explaining the workability of the GCC currency bloc.
7. Concluding Remarks We have argued that from the standpoint of traditional OCA analysis, the GCC countries are a major anomaly. They have successfully maintained pegs among themselves for two decades despite their failure to meet a number of important OCA criteria. Of course traditional OCA analysis has not been free of criticism, the strongest coming from new classical macro economists who assume wage and price flexibility that eliminates most of the domestic political cost associated with giving up independent discretionary monetary policy. However, the GCC economies can hardly be described as closely approximating the assumptions of the new classical critique of OCA analysis. 13 We see little reason to believe that their economies are so flexible that there is no need for discretionary macroeconomic policy14. Nor have endogenous OCA considerations stimulated great increases in intraregional trade and business cycle correlations. However, until recently, there had been a substantial convergence of inflation rates. And this was achieved without generating major domestic unemployment problems. 13
For references to and assessments of these new classical critiques, see De Grauwe, Tavlas(1993) and Willett and Wihlborg (1999). 14 For the opposite view see Buiter (2007). 22
We suggest that another type of mechanism has been at work to largely shield domestic workers from the effects of aggregate economic fluctuations in the face of considerable domestic wage stickiness and low labor mobility. This mechanism is government protection of the jobs of most citizens. This in turn has been feasible because of the economic and oil wealth and small private sectors of the GCC countries combined with the heavy use of foreign workers to provide flexibility. Such conditions were not envisioned in traditional OCA analysis and likely apply to few economies outside of the GCC. Thus, in our judgment, the anomaly of the GCC does not undercut the usefulness of OCA analysis for most countries. Furthermore, traditional OCA considerations are likely to become more relevant for at least some members of the GCC in the future. Some GCC countries are fast approaching the end of their hydro carbon reserves while at the same time the size of private sectors are growing. As the GCC economies become more market oriented, the power of the government employment cushion will decline and already we are beginning to see a substantial pick up in the levels and differentials in inflation rates across the GCC countries. This indeed was the catalyst for Kuwait‟s recent switch in exchange rate policy. In response to the current global financial crisis and recession, GCC countries are beginning to adopt simulative fiscal policies. As the investment funds warn, past success does not guarantee future performance.
23
References
Abed, G. et al. (2003), ‘The GCC Monetary Union: Some considerations for the Exchange Rate Regime’ IMF Working Paper – WP/03/66, April. Abu-Qarn, Aamer S. and Suleiman Abu-Bader(2008), ‘On the Optimality of a GCC Monetary Union: Structural VAR, Common Trends, and Common Cycles Evidence’ The World Economy, 31, 5, 612-630. Al-Bassam, K.. (2002), ‘The Gulf Cooperation Council Monetary Union: a Bahraini Perspective’ Bank of International Settlements, 17. Al-Jasser, M. and A. Abdulrahman (2002), ‘A Common Currency for the Gulf Region’ Bank of International Settlements, 17. Buiter, Willem H. “Economic, Political, and Institutional Prerequisites for Monetary Union among the Members of the Gulf Cooperation Council “ A Paper Presented in Conference in Dubai, UAE, March 20-21, 2006. Darvas, Z. and G. Szapary (2004), ‘Business Cycle Synchronization in the Enlarged EU: Comovements in the (Soon-To-Be) New and Old Members’ Magyar Nemzeti Bank Working Paper -2004-1 De Grauwe, Paul (1994), The Economics of Monetary Integration, 32nd. Revised Edition, New York: Oxford University Press. Fasano, U. and Q. Wang (2001), ‘Testing the Relationship Between Government Spending and Revenue: Evidence from GCC countrie’ IMF Working Paper – WP/02/195, December. Fasano, U. and Q. Wang, (2002), ‘Fiscal Expenditure Policy and Non-Oil Economic Growth: Evidence from the GCC countries’, IMF Working Paper – WP/02/201, November. Frankel, J. and A. Rose (1998), ‘The Endogeneity of the Optimum Currency Area Criteria’ The Economic Journal, 108, 449, July. Gulf Research Center (2005), ‘GCC Integration – More Rhetorical than Real?’ www.gulfinthemedia.com June 27th.
24
Gulf Research Center (2005), ‘Co-operation Council for the Arab States of the Gulf (GCC Charter)’ www.gulfinthemedia.com Sept. 15th. Jadresic, E. (2002), ‘On a Common Currency for the GCC Countries’ IMF Policy Discussion Paper – PDP/02/12, December. Laabas, B. and I. Limam (2002), ‘Are GCC Countries Ready for Currency Union?’ Arab Planning Institute – Kuwait, April. Mongelli, F. P. (2002), ‘New views on the Optimum Currency Area Theory: What is EMU Telling Us?’ European Central Bank – Working Paper Series, 138, April. Okogu, B. (2003), ‘The Middle East and North Africa in a Changing Oil Market’ IMF Publications. Washington: International Monetary Fund. Sept. 5. . Popescu, M. and S. Mustafa (2001) ‘The Gulf Monetary Unification: Opportunities and Challenges’ Arab Bank Review, 3, 1, April. Rutledge, E. (2005), ‘Will Record Oil Prices Help Economic Diversification?’ www.gulfinthemedia.com July 28th. Rutledge, E. (2006), ‘The dollar declines, while the euro shines’ www.gulfinthemedia.com May 19th. Sturm, Michael and Siegfried Nikolaus (2005) ‘Regional Monetary Integration in the Member States of the Gulf Cooperation Council’(June 2005). ECB Occasional Paper No. 31, June. Available at SSRN: http://ssrn.com/abstract=752091 Tavlas, G. S. (1993), ‘The New Theory of Optimum Currency Areas’ The World Economy, 16, 6, 663-685. Tavlas, G. S. (1994), ‘Theory of Monetary Integration’ Open Economies Review, March 5, 2, 211-230. Tower, E. and T. Willett (1976), ‘The Theory of Optimum Currency Areas and Exchange Rate Flexibility’, Special Papers in International Economics, Princenton University, 11. Vaubel, R. (1976), ‘Real Exchange-Rate Changes in the European Community: The Empirical Evidence and Its Implications for European Currency Unification, Weltwirtschaftliches Archiv, 112, 3, 429-70. Willett, T. and C. Wihlborg (1999), ‘The Relevance of the Optimum Currency Area Approach for Exchange Rate Policies in Emerging Market 25
Economies’, in R. J. Sweeney, C. G. Wihlborg, and T. D. Willett (eds), Exchange-Rate Policies for Emerging Market Economies (Boulder, CO: Westview Press). Willett, T. (2003), ‘The OCA Approach to Exchange Rate Regime’, in D. Salvatore, J. Dean, and T.Willett (eds), The Dollarization Debate (Oxford: Oxford University Press). Willett, T. (2003),’Optimum Currency Area and political Economy Approaches to Exchange Rate Regimes: Towards a Framework for Integration’ Current Politics and Economics in Europe, 17, 1, 25-52. Willett, T., O. Permpoon, and C. Wihlborg (2007), ‘Endogenous OCA analysis and the Early Euro Experience’, Claremont working paper.
26
The GCC’s Fixed Exchange Rate: A Major Anomaly for OCA Analysis
Thomas D. Willett* Khalfan M. Al-Barwani** Sherine M. El Hag*** School of Politics and Economics Claremont Graduate University California
*Corresponding Author Director, Claremont Institute of Economic Policy Studies and Horton Professor of Economics Claremont Colleges [email protected] **Part time lecturer at Sultan Qaboos University *** Lecturer California State University, Fullerton [email protected]
Abstract: From the stand point of traditional OCA analysis, the GCC countries are a major anomaly. They successfully maintained pegged exchange rates among themselves for two decades while failing to meet a number of major OCA criteria. Contrary to the optimistic predictions of some of the most enthusiastic advocates of endogenous OCA analysis, intra-regional trade has remained low and there is little synchronization of their business cycles or harmonization of their fiscal positions. We argue that their success in operating a fixed rate regime has been due to a combination of high oil revenue, the heavy use of foreign workers, and the limited extent to which their economies rely on market forces. These are conditions that are not met by most other countries.
1. Introduction The Gulf Cooperation Council (the GCC) countries have pledged to form a monetary and currency union1. Toward this end, the six GCC countries officially pegged their respective currencies both in terms of de jure and de facto to the US dollar in 2003. By 1986, all the GCC member states, except for Kuwait 2, already had de facto pegs to the US dollar. The original date set for the union was 2010, but this deadline was recently extended. The official explanation for the delay was that more time was needed to complete the development of a common monetary and regulatory framework. There had been previous signs that the 2010 goal was unlikely to be met, with Oman announcing in 2006 a need for it to delay. However, there is still strong support for eventual monetary union among many of the GCC‟s national leaders and it remains the likeliest candidate for the creation of the next sizeable currency union. One of the most surprising characteristics of the GCC economies is that from 1981 until 2007 when Kuwait revalued its dinar and switched from a dollar to a basket peg, there were no major changes in intra group exchange rates, and the Kuwait revaluation was itself quite modest. Furthermore, despite the absence of capital controls, there were few instances of large speculative pressures until the current global crisis. 1
The GCC countries consist of Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates (UAE). 2
Until 2003, the Kuwaiti exchange rate regimes differed slightly from the rest of the GCC countries since the Kuwaiti Dinar was pegged to a basket of currencies of its main trading and financial partners. While the composition of the basket was not disclosed, the US dollar was believed to be the dominant weight since minor fluctuations occurred visà-vis the US dollar and subsequently the fluctuations vis-à-vis other GCC currencies was also fairly small (Sturm and Siegfried, 2005). 1
The reason for the intra group exchange rate stability likely resulted primarily from the GCC countries‟ common desires to peg to the dollar because of the pricing of oil in dollars3. What is surprising is that it was so successful despite the GCC economies failing to meet a number of the major criteria for being an optimal currency area (OCA). The answer to this puzzle we suggest lies in several unusual characteristics of the GCC economies that differ substantially from the standard assumptions implicitly underlying traditional OCA analysis. In section 2, we provide historical and institutional background on the GCC. In section 3, we briefly review how the GCC economies fit with the major OCA criteria. We find that in general they fail to meet most of these criteria. For example, while they all are highly open economies, they have relatively little trade with each other and this is what is relevant for the economic costs and benefits of forming a currency area. In section 4, we provide a suggested explanation for why the GCC‟s pegged rate regime worked so well for so long despite its failure to meet the traditional OCA criteria. In sections 5 and 6, we consider other possible explanations. In section 5, we critique Willem Buiter‟s recent critique of OCA analysis and in section 6, we consider the possibility suggested by the recently developed endogenous OCA analysis that the fixing of GCC exchange rate unleashed powerful forces to move the GCC countries much closer to meeting the OCA criteria. We do not find evidence of strong endogenous effects on trade flows, business degree synchronization, or fiscal policy coordination. Section 7 offers concluding comments. We suggest that the success of exchange rate fixity within the GCC provides
3
We are indebted to a referee for suggesting this point. 2
little promise that the macroeconomic costs of fixed exchange rate would be low for countries that do not share these characteristics.
2. Historical and Institutional Background In May 1981, the Arab Gulf region countries: Bahrain, Kuwait, Oman, Qatar, Saudi Arabia and the United Arab Emirates, formed the Cooperation Council for the Arab States of the Gulf (GCC countries). The main objectives of the GCC include strengthening the economic cooperation among Member Countries, harmonizing their economic and financial policies, promoting the private sector, and achieving a common market and monetary union. There are a number of common characteristics among the GCC counties. These include heavy reliance on hydro carbon resources for their domestic and external economies, geographical proximity, small private sectors and dominant public sectors, extensive reliance on foreign labor force in the private sector, common language and culture, and small populations except for Saudi Arabia where about 70% of the estimated 35 million combined population of the GCC reside. The GCC charter, known as “Charter of the Cooperation Council”, established the main bodies of the GCC, which consist of the Supreme Council, the Ministerial Council and the Secretariat General. The Charter also laid down the compositions of these bodies and their functions. They include the intergovernmental decision-making and consultative bodies with the Supreme Council and the Ministerial Council at its core, the Secretariat General as the main supranational institution, plus a number of specialized regional agencies.
3
The Supreme Council is the highest authority of the governmental bodies where all main decisions are made. This Council consists of the Heads of State of the six GCC member states and its presidency is rotated on annual basis. The Supreme Council, which meets twice a year, has the task of providing policy direction, reviewing the reports and recommendations submitted by other bodies and appointing the Secretary General of the GCC. Foreign Ministers or other ministers of the member states form the Ministerial Council. This Council is tasked with policy proposals, preparing recommendations, and conducting studies in all concerned areas. A number of specialized committees also work under the Ministerial Council. The most important committee with regard to economic integration is the Committee for Financial and Economic Cooperation, which includes the Ministers of Finance and Economies. The various ministerial committees direct a number of sub-committees to conduct studies, write proposals and coordinate national policies at a technical level. Resolutions passed by the Supreme Council as well as the Ministerial Council are on the basis of unanimity for major issues and by majority rule for procedural matters. Cooperation among the GCC central banks takes place under the Monetary Agencies and Central Bank Governor‟s Committee. This committee meets twice a year and reports to the Committee for Financial and Economic Cooperation and its subcommittees, which consist of the Banking Supervision Committee, the Payment Systems Committee, the Monetary Union Committee, and the Training Committee.
4
According to the Charter, the Secretariat General works independently and for the common interests of the member states, and unlike the two other regional bodies, it does not consist of representatives of the member states. Its headquarters is located in Riyadh, Saudi Arabia and the Secretary General, who heads this institution, serves a three-yearterm subject to one renewal. The functions of the Secretariat General include administrative preparation, follow-up of the meetings of the intergovernmental bodies, preparation of studies and reports on matters pertaining to the cooperation objectives, follow up the implementations of resolutions and recommendations, and draft resolution on common legislation. The Secretariat also acts as a representative of the six member states in certain international forums. Additionally, this body has been assigned the task of settling disputes regarding the implementation of the GCC Economic Agreement and its ratified decisions, including claims raised by private parities. In the event the Secretariat General fails to settle a dispute, it can refer the matters to the GCC Commercial Arbitration Center, a judicial body. The six member states have also set up some specialized agencies to deal with issues related to patent rights, commercial disputes, and product standardization. In November 1981, the GCC established an economic agreement to set the stage for full economic integration (Laabas and Liman, 2002). In 1982, the goal to create a unified currency was declared by the six states. In 1983, a free trade zone was formed as a first step toward their economic integration. However, it took the GCC states sixteen years to reach a custom union agreement in 1999. In 2000, the GCC countries agreed to adopt a common exchange rate peg as a step toward the realization of a common currency in 2010. In 2001, an accord on a joint custom tariff of five percent was reached
5
and implemented two years later. In 2002, the US dollar was selected as the intermediate peg for the six member states‟ currencies before the proposed unified currency comes into effect. Officially, the six GCC member states began to peg their respective currencies to the US dollar in 2003 as a step towards adopting a common currency projected to take place in 2010. Unofficially, this policy was adopted much earlier. In fact, by 1986, all of the GCC member states, except for Kuwait 4 had initiated their de facto pegs to the US dollar. In early 2004, the GCC heads of central banks agreed in principle on key convergence criteria for entry into the common currency. These included the size of budget deficits, inflation rates, interest rates, foreign reserves and the ratio of public debt to gross domestic product. There is not general agreement on the primary reasons for the GCC‟s pursuit of a monetary union. While some have suggested economic explanations based on OCA considerations ( see, for example, Al-Jasser and Al-Hamdy(2003), others have suggested that both the economic costs and benefits would be modest, see, for example, Sturm and Siegfried (2005) and still others have concluded that important OCA conditions are not met (Abu-Qarn and Abu-Bader (2008). In general, we find that while economic arguments are often used, the actual formation of currency areas is usually based more on geopolitical considerations.
4
Until 2003, the Kuwaiti exchange rate regimes differed slightly from the rest of the GCC countries in the sense that the Kuwaiti Dinar was pegged to a basket of currencies of its main trading and financial partners. While the composition of the basket was not disclosed, the US dollar was known to be the dominant weight since only minor fluctuations occurred vis-à-vis the US dollar and subsequently the fluctuations vis-à-vis other GCC currencies were also fairly small (Sturm and Siegfried, 2005). 6
This was the case in Europe with the creation of the Euro and was likely important also with the GCC5. Also likely is the dynamic observed in Europe of pushing from one stage of economic integration to another. While the experience of the North America Free Trade agreement among Canada, Mexico, and the United States shows that countries can take major steps to increase trade integration without contemplating monetary integration, in Europe there were strong political pressures to continue on to monetary union. Since there is relatively little trade among GCC countries, it seems likely that there was much less pressure for monetary integration from economic interest groups then there was in Europe. Since integration efforts were begun the GCC countries made considerable progress on reducing inflation rates and this was maintained until quite recently. This resulted in considerable convergence of inflation rates. Rates of economic growth have shown considerable divergence, however. For example, over the period 1975 to 2003, average growth rates range from at high of almost 7 percent in Oman to a low of below 1 percent in Kuwait. The largest country, Saudi Arabia, averaged 2.4 percent (Abu-Qarn and AbuBader,2008).
3. The GCC Anomaly and a Suggested Explanation The theory of optimum currency areas lays out the factors that explain the costs and benefits of countries adopting hard fixed (i.e. joining a currency area) versus flexible rates. The major cost of joining a currency area is giving up independent monetary and exchange rate policies. This cost is low if the domestic economy is quite flexible and 5
On Europe see the analysis and references in Willett(2003 ). On the role of geopolitical considerations within the GCC see Popescu and Mustafa(2001). 7
could be high if the economy is rigid. Furthermore, if the economy is highly open, exchange rate adjustments lose their effectiveness, while the importance of the external sector grows relative to the internal one. Thus, large closed economies with domestic rigidities are ideal candidates for flexible rates while small open economies with flexible domestic labor and product markets are ideal candidate for fixed rates. This assumes, however, that the country in question has a stable partner or group of countries to which to fix its exchange rate. Thus, patterns of shocks and correlation of macro economic conditions could be important as well. 6 In recent years, a major research industry has developed applying OCA criteria to various countries or group of countries. Of necessity, most of these empirical studies have looked at attributes of countries under various forms of flexible rates and tried to predict how well the fixing of exchange rates would work.
The case of the GCC
countries, however, provides a natural experiment of how well they operated under fixed rates. While not all the major OCA criteria point in the same direction (for a summary of previous studies of the GCC see Abu-Qarn and Abu-Bader(2008), our reading is that the most relevant OCA conditions are far from met. As Abu-Qarn and Abu-Bader conclude, "In line with most of the previous studies, the three methods employed here provide no evidence of the readiness of the GCC members to establish a lasting and well functioning currency union." (2008,p 629). What is surprising is that despite failing to meet many of these criteria, the GCC countries have successfully maintained a regime of fixed exchange rates for two decades without the widespread use of controls. How can this be?
6
On the development of OCA theory pioneered by Mundell, Mckinnon, and Kenen and surveys of recent developments see Tavlas (1993) and Willett (2003). 8
At first glance, all of the GCC countries except Saudi Arabia meet the major OCA criteria of being small open economies. All the countries except Saudi Arabia have ratios of imports to GDP of greater than 30%. What is relevant for currency formation, however, is how open the prospective members are to trade with each other and on this score the GCC countries score quite low. Only Oman had more than 10% of its exports go to other GCC countries. In 2003, the average ratio of intra group to total exports was only 5 per cent. Low intra- area trade was the reason why the three small Baltic States (Estonia, Latvia, and Lithuania) did not seriously consider forming a common currency area after the break up of the former Soviet Union. While there is not enough data available to do formal analysis, it is generally judged that levels of factor mobility by GCC citizens are low, as is the degree of wage flexibility7. However, GCC labor markets are “highly unusual” to use William Buiter‟s (2007) phrase. They have more flexibility than these conditions suggest because of the heavy use of expatriate workers. These expatriates work primarily in the private sector, while nationals work primarily in the public sector which is characterized by wage rigidity, immobility, and inflexibility. Foreign labor constitutes about 55% of total employment in Bahrain and more than 80% in the UAE, Kuwait and Qatar 8. This combination of high levels of protected jobs for the locals and large movements of foreign workers cushion the domestic effects of macroeconomic fluctuations despite the domestic rigidities. While this process has provided considerable adjustment in the past, Willem Buiter (2007) questions whether it will continue to operate 7
For more detailed analysis of the GCC economies, see Jadresic, 2002, Laabas and Liman (2002), and Sturn and Siegfried (2005) 8 The expatriate labor is usually from India, Philippines and Indonesia, other MiddleEastern nations, Europe and USA. 9
as well in the future. The labor mobility among the expatriates “can be a substitute for the removal of legal and administrative obstacles to cross-border labor mobility by resident workers, native and expatriate, which would [be] the standard way to use labor mobility as a substitute for exchange rate flexibility. It is unclear how long the current migration paradigm of the GCC countries can survive. At some point the „guest workers‟ will demand greater rights and will no longer be willing to accept passively to act as the residual in the GCC labor markets” ( p27) .
4. The Anomaly Explained Traditional OCA analysis was developed within the context of economies with large private sectors and only intra- area labor movements. In such economies, substantial labor mobility and/or wage and price flexibility is needed to cushion the efforts of macroeconomic fluctuations under fixed exchange rates since the use of independent monetary policies is taken away. The GCC countries have been able to compensate for the lack of such domestic flexibility to a considerable degree through the mobility of expatriates described above. Combined with limited need for adjustment by nationals in the private sector, high levels of revenue from oil sales( see table 1) have given governments the capacity to effectively protect the jobs of domestic workers without severe fiscal strains.
10
Table 1 Oil Dependency of the GCC and its Member Countries S. ARABIA
UAE
GCC
56.80
34.90
28.10
36.00
78.40
64.20
80.60
75.30
79.30
64.50
34.50
65.50
38.80
66.70
BAHRAIN
KUWAIT
OMAN
QATAR
GDP*
25.70
45.90
43.10
Govt. Revenue**
73.00
91.50
Exports***
66.70
83.80
Oil as Share of
Sources: European Central Bank (ECB), Arab Monetary Fund (AMF), GCC central banks, Internationa Mnetary Fund (IMF), and Institute of International Finance (IIF) * Oil and gas sector's share of GDP as a % in 2001 ** Oil revenue/total government revenue as a % (include gas revenue for Bahrain) in 2000 *** Oil and oil products' share of total exports as a % in 2004
This insulation of nationals from the effects of macroeconomic fluctuations has allowed the GCC countries to achieve considerable harmonization in their rates of inflation until recently (see chart 1 below). This was done without generating the types of strong domestic unemployment pressures that would be generated by giving up countercyclical monetary policies in the types of Keynesian economies analyzed in the initial OCA contributions.
11
Chart 1: GCC Inflation Rates 1980 – 2004
Data Source: WDI Note that Qatar data were unavailable for the period under review.
Of course even with harmonized inflation rates, the GCC countries‟ high levels of export instability generated by their heavy reliance on petroleum exports could generate pressures on intra group exchange rates. With fairly similar export structures, however, we would expect the symmetry of the major shocks to their export revenues to serve to reduce the tensions generated on intra group exchange rates. Surprisingly, however, AbuQarn and Abu-Bader(2008) find that the supply shocks to the GCC countries are asymmetric, while the demand shocks are more symmetric. They suggest the possible explanation that for the oil producers shifts in the international price of oil act as both supply and demand shocks. 12
This doesn‟t help with the GCC countries‟ abilities to maintain their external peg to the dollar, however. Here their high levels of oil revenues have allowed them to avoid much of the domestic macroeconomic adjustment that would have been needed in the absence of substantial scope for financing deficits. Again, standard OCA analysis assumes that countries must adjust to external imbalances that are not fairly quickly mean reverting. The GCC countries oil wealth substantially relaxed this constraint.
5. A Critique of Buiter’s Critique of OCA Analysis Our argument for the limited relevance of traditional OCA analysis for the GCC countries is quite different from the analysis by Willem Buiter (2007)9. He argues the much stronger proposition that “the theory of optimal currency areas…is almost entirely useless as a guide to the choice of currency regime in modern economies” (p15). With this generalization, we strongly disagree. Buiter argues that there are two key failures of the OCA approach. One is what he calls “chronic confusion between transitory nominal rigidities and permanent real rigidities” (p). We share Buiter‟s view that there is no favorable long run inflationunemployment trade-off such as was assumed in early Keynesian analysis ,but this needn‟t imply that nominal rigidities are not of sufficient importance to discretionary monetary policy to reduce the costs of adjusting to some types of economic shocks. Buiter‟s second argument is that “conventional OCA literature was designed for a world without endogenous capital mobility” (p16). His argument that capital flows can in some circumstances amplify disturbances is well taken. 9
While Buiter favors a GCC monetary union on economic grounds, he is skeptical that necessary political conditions are met. 13
However, there are also circumstances where capital mobility can dampen the effects of shocks. That high capital mobility can increase or reduce macroeconomic stability depending on both nature of the shock and the exchange rate regime is a basic result of Mundell-Fleming analysis (see Tower and Willett 1976). Buiter‟s real point is that destabilizing speculation can make the foreign exchange market an exogenous source of disturbance. While an unusual view for a new classic macro economist to hold, destabilizing speculation is certainly possible. This raises an important issue, but doesn‟t call for a full scale rejection of OCA analysis. One final comment on Buiter‟s analysis is in order. He describes all of the GCC countries as being small, open economies including Saudi Arabia. He uses a different criterion for openness than does conventional OCA analysis. He focuses on the ability of countries to influence their terms of trade. This is not necessary for exchange rate adjustment to be effective, however. The type of openness relevant for this is that there be enough non-traded goods and services in the economy so that a change in the nominal exchange rate can lead to a meaningful change in the relative price of traded and nontraded goods.
6. Endogenous OCA Analysis: Did the establishment of pegged exchange rates lead to a greater convergence in business cycle conditions? Another possible explanation for the success of the GCC fixed exchange rate regime comes from endogenous OCA theory that emphasizes that the act of fixing exchange rates may change the behavior of OCA criteria in ways that make fixed exchange rates more workable.
14
Fixing exchange rates should increase intra-regional trade and likely also increase the similarities of macroeconomic fluctuations, thus reducing the divergences among optimal national monetary policies. The creation of the euro area has been consistent with both of these predictions, although the concomitant increases in trade and business cycle correlations with other Europeans countries make it difficult to judge how much of these changes have been due to the creation of the euro per se10. For the GCC countries, however, the increase in intra regional trade over the fixed rate period has been modest. Table 2 describes the evolution of intra GCC trade.
10
See Willett, Permpoon, and Wihlborg (2007). Of course, the creation of a common currency would be expected to have stronger endogenous effects than the GCC‟s pegged rate regime. 15
Table 2: Intra GCC Exports & Imports as Shares of Total Exports and Imports from 1986 to 2003 FIVE-YEAR AVERAGES
19841988
19891993
1994-1998
1999-2003
GCC EXP % TOTAL EXP
GCC EXP % TOTAL EXP
GCC EXP % TOTAL EXP
GCC EXP % TOTAL EXP
15.621 2.584 5.252 5.480 5.891 4.526
6.640 1.997 8.180 6.536 6.751 5.275
8.837 1.613 12.543 5.400 7.437 7.066
7.002 1.525 11.984 4.937 5.195 6.457
1984-1988
1989-1993
1994-1998
1999-2003
GCC IMP % TOTAL IMP
GCC IMP % TOTAL IMP
GCC IMP % TOTAL IMP
GCC IMP % TOTAL IMP
44.169 4.234 22.567 9.983 1.881 9.371
41.631 4.443 26.988 11.129 1.907 10.066
37.244 9.638 27.429 13.295 3.050 9.440
25.610 11.629 32.205 14.970 2.951 8.340
COUNTRY BAHRAIN KUWAIT OMAN QATAR SAUDI ARAB UAE FIVE-YEAR AVERAGES
COUNTRY BAHRAIN KUWAIT OMAN QATAR SAUDI ARABIA UAE
Data Source: IMF Direction of Trade Statistics Intra-GCC Export as % of Total Export indexes are computed by the authors
To investigate the degree of synchronization of macroeconomic fluctuations, we computed the correlations among the first differences in the logs of real GDP over one, two, and three year time intervals. The recent literature has generally assumed that high correlations increase the case for a currency union since this would give rise to smaller differences in optimal discretionary monetary policies. Thus the constraint of adopting a
16
common policy would be less costly. We are doubtful, however, that monetary policy can normally be fine-tuned sufficiently to respond effectively to one year fluctuations. The combined lags in recognition and effects of changes in monetary policy on the economy are likely to be too long. So despite the common tendency in the literature to focus only on one year correlations, we believe that correlations over longer time periods of two to three years are more relevant for analyzing the scope of effective discretionary monetary policy responses. Indeed, it can be argued that over horizons of a year or less, low correlations would be superior as these would fit the automatic stabilization criteria emphasized in some of the OCA literature11. Table 3 shows the correlation matrices for 1970 through 2004.
11
See Tower and Willett ( 1976) 17
Table 3 GCC Growth Rate Correlation Matrices Table 3a: GCC One-year Growth Rate Correlations BAHRAIN
KUWAIT
OMAN
QATAR
SAUDI ARABIA
UAE
BAHRAIN
1.000
KUWAIT
0.022
1.000
OMAN
0.192
-0.130
1.000
QATAR
-0.255
0.248
-0.357
SAUDI ARABIA
0.438
-0.166
-0.159
0.244
1.000
UAE
0.623
-0.202
-0.057
-0.149
0.692
1.000
QATAR
SAUDI ARABIA
UAE
1.000
Table 3b: GCC Two-year Growth Rate Correlations BAHRAIN
KUWAIT
OMAN
BAHRAIN
1.000
KUWAIT
-0.012
1.000
OMAN
0.122
-0.213
1.000
QATAR
-0.445
0.187
-0.492
SAUDI ARABIA
0.496
-0.206
-0.247
0.226
1.000
UAE
0.716
-0.315
-0.042
-0.168
0.791
1.000
QATAR
SAUDI ARABIA
UAE
1.000
Table 3C: GCC Three-year Growth Rate Correlations BAHRAIN
KUWAIT
OMAN
BAHRAIN
1.000
KUWAIT
-0.062
1.000
OMAN
0.178
-0.288
1.000
QATAR
-0.384
0.131
-0.645
SAUDI ARABIA
0.625
-0.236
-0.267
0.148
1.000
UAE
0.741
-0.357
0.009
-0.126
0.869
1.000
1.000
Data source: IMF/IFS from 1970 to 2004 Computed by the Authors
The most striking feature of these calculations is the overall low magnitudes of most of the correlation coefficients. Moreover, a number of the pairwise correlation coefficients have negative signs. The highest correlation coefficients are between Saudi Arabia and the UAE and this is consistent in all three matrices. The other high correlation coefficients observed are those between Bahrain and the UAE with the highest magnitude of .74 for three-year growth rates and lowest of .62 for one year growth rates and between Bahrain and Saudi Arabia with the highest magnitude of 0.625 in the three-year growth
18
rate matrix and the lowest under the one-year growth rate matrix at 0.4 12. Across all three matrices there are consistent negative correlation coefficients between Oman and Qatar, Bahrain and Qatar, and of smaller magnitude, between Oman and Saudi Arabia. The GCC countries‟ average correlation coefficients for one-year growth rates are slightly higher than for the two and three-year averages. This conflicts with the ideal pattern which would display low or negative short run correlations for automatic stabilization and high long term correlations so that there would be less difference in optimal discretionary monetary policies response across countries. Too much should not be made of these correlation tables, however, since the correlations can vary substantially over time and not always along a steady path. This is illustrated in chart 2 which displays ten year rolling window calculations of the three year growth correlations of the five smaller GCC countries against Saudi Arabia. These give the ten year moving averages of these correlations. If there is any clear pattern, it is for both the high positive and high negative correlations displayed in the 1980s to converge toward lower correlations. In a number of cases, the correlations change signs in the latter part of the period. The drop in positive correlations with Saudi Arabia is especially strong for Oman and the UAE. Possible explanations include differences in the extent and speed of fiscal response to shocks from one country to another, and in the pace of economic diversifications.
12
The Bahraini economy has for years depended on Saudi grants of crude oil that it refines, consumes domestically and exports. 19
Chart 2: Three-Year GCC RGDP Growth, Ten-Year Rolling Window Correlations Three-Year Growth Rate: Ten-Year Busines Cycle Rolling Window Correlation, Between Saudi Arabia and Other GCC Countries 1970-2004 Correlations
1.20 1.00 0.80 0.60 0.40 0.20 0.00 -0.20 -0.40 -0.60 -0.80
BAHRAIN
KUWAIT
OMAN
QATAR
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
-1.00
UAE
Data source: IMF/IFS from 1970 to 2004
Given the similarity of major shocks through the oil markets, the generally low levels of business cycle correlations among the GCC countries is surprising. A part of the explanation is that while monetary policies have led to substantial convergence of inflation rates, fiscal balances have varied widely among the countries. Table 4 presents data on fiscal imbalances as a percent of GDP and changes in the ratio of government expenditure to revenue for the six GCC countries over various time periods and business cycles. Sturm and Siegfried (2005) find that while the underlying components of the budget deficits- government revenue and expenditures- exhibit high degrees of comovements, changes in the ratio of expenditure to revenue do not follow a similar pattern. Using a co-integration and error- correction modeling framework, Fasano and Wang (2001) find that government spending in the GCC countries is driven by revenue from oil, suggesting a pro-cyclical expenditure policy. However, the timing and extent of spending changes differs among countries and this results in different patterns shown in
20
table 4. Thus fiscal policy coordination cannot be used to explain the success in maintaining the GCC countries‟ internal pegs. Table 4: GCC Fiscal Policy Indicators Average form to
1975 1980
1980 1986
1986 1990
1990 1994
1994 1996
1996 1998
1998 2000
2000 2002
Deficit (-) or Surplus as % of GDP
-0.30
0.71
-4.92
-3.86
-3.65
-4.22
-2.73
n.a.
Change in expenditure/revenue*
0.90
-7.60
1.30
-1.50
0.40
-0.30
0.20
n.a.
Deficit (-) or Surplus as % of GDP
41.48
5.99
n.a.
n.a.
-3.40
1.77
4.45
19.34
Change in expenditure/revenue*
0.70
-0.90
-15.80
-2.00
0.20
-0.70
0.00
n.a.
Deficit (-) or Surplus as % of GDP
-1.59
-8.22
-9.24
-7.98
-7.67
-3.95
-6.52
n.a.
Change in expenditure/revenue*
0.50
7.20
0.00
-1.70
0.00
0.10
0.20
n.a.
Bahrain
Kuwait
Oman
Qatar Deficit (-) or Surplus as % of GDP
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
Change in expenditure/revenue*
0.60
0.10
0.10
-0.90
0.80
-0.20
0.10
n.a.
Deficit (-) or Surplus as % of GDP
n.a.
n.a.
n.a.
n.a.
n.a.
-3.06
-4.85
n.a.
Change in expenditure/revenue*
0.90
0.50
0.50
1.20
0.40
0.20
0.40
n.a.
Deficit (-) or Surplus as % of GDP
0.64
n.a.
n.a.
0.10
-0.14
0.19
n.a.
n.a.
Change in expenditure/revenue*
1.00
0.20
0.50
0.20
2.20
1.50
0.30
n.a.
234.80
-65.00
62.20
-30.60
27.70
-35.80
116.00
n.a.
Saudi Arabia
UAE
Change in average crude oil prices** Data source: IMF/IFS. Indices computed by the author *Indices provided by Fasano and Wang (2001)
** By Fasano and Wang (2001) based on average nominal price of UK Brent, Dubai, and West Texas Intermediate
Another factor often posited to help the workings of a currency area is financial integration which in turn would be encouraged by the establishment of fixed exchange rates. This doesn‟t appear to have been a major factor within the GCC countries. Despite the removal of most restriction on capital flows, there has been relatively little movement of financial capital within the GCC. The vast majority of capital flows are with outside countries and a substantial intra-GCC banking market is yet to appear (see Sturn and
21
Siegfried, 2005). Factors explaining the low movement of capital among the GCC countries include bureaucratic barriers and limited investment opportunities. The GCC financial markets are generally still in their infancy. In summary, there is little support for the hypothesis that strong endogenous responses to pegged exchange rates have been a major factor in explaining the workability of the GCC currency bloc.
7. Concluding Remarks We have argued that from the standpoint of traditional OCA analysis, the GCC countries are a major anomaly. They have successfully maintained pegs among themselves for two decades despite their failure to meet a number of important OCA criteria. Of course traditional OCA analysis has not been free of criticism, the strongest coming from new classical macro economists who assume wage and price flexibility that eliminates most of the domestic political cost associated with giving up independent discretionary monetary policy. However, the GCC economies can hardly be described as closely approximating the assumptions of the new classical critique of OCA analysis. 13 We see little reason to believe that their economies are so flexible that there is no need for discretionary macroeconomic policy14. Nor have endogenous OCA considerations stimulated great increases in intraregional trade and business cycle correlations. However, until recently, there had been a substantial convergence of inflation rates. And this was achieved without generating major domestic unemployment problems. 13
For references to and assessments of these new classical critiques, see De Grauwe, Tavlas(1993) and Willett and Wihlborg (1999). 14 For the opposite view see Buiter (2007). 22
We suggest that another type of mechanism has been at work to largely shield domestic workers from the effects of aggregate economic fluctuations in the face of considerable domestic wage stickiness and low labor mobility. This mechanism is government protection of the jobs of most citizens. This in turn has been feasible because of the economic and oil wealth and small private sectors of the GCC countries combined with the heavy use of foreign workers to provide flexibility. Such conditions were not envisioned in traditional OCA analysis and likely apply to few economies outside of the GCC. Thus, in our judgment, the anomaly of the GCC does not undercut the usefulness of OCA analysis for most countries. Furthermore, traditional OCA considerations are likely to become more relevant for at least some members of the GCC in the future. Some GCC countries are fast approaching the end of their hydro carbon reserves while at the same time the size of private sectors are growing. As the GCC economies become more market oriented, the power of the government employment cushion will decline and already we are beginning to see a substantial pick up in the levels and differentials in inflation rates across the GCC countries. This indeed was the catalyst for Kuwait‟s recent switch in exchange rate policy. In response to the current global financial crisis and recession, GCC countries are beginning to adopt simulative fiscal policies. As the investment funds warn, past success does not guarantee future performance.
23
References
Abed, G. et al. (2003), ‘The GCC Monetary Union: Some considerations for the Exchange Rate Regime’ IMF Working Paper – WP/03/66, April. Abu-Qarn, Aamer S. and Suleiman Abu-Bader(2008), ‘On the Optimality of a GCC Monetary Union: Structural VAR, Common Trends, and Common Cycles Evidence’ The World Economy, 31, 5, 612-630. Al-Bassam, K.. (2002), ‘The Gulf Cooperation Council Monetary Union: a Bahraini Perspective’ Bank of International Settlements, 17. Al-Jasser, M. and A. Abdulrahman (2002), ‘A Common Currency for the Gulf Region’ Bank of International Settlements, 17. Buiter, Willem H. “Economic, Political, and Institutional Prerequisites for Monetary Union among the Members of the Gulf Cooperation Council “ A Paper Presented in Conference in Dubai, UAE, March 20-21, 2006. Darvas, Z. and G. Szapary (2004), ‘Business Cycle Synchronization in the Enlarged EU: Comovements in the (Soon-To-Be) New and Old Members’ Magyar Nemzeti Bank Working Paper -2004-1 De Grauwe, Paul (1994), The Economics of Monetary Integration, 32nd. Revised Edition, New York: Oxford University Press. Fasano, U. and Q. Wang (2001), ‘Testing the Relationship Between Government Spending and Revenue: Evidence from GCC countrie’ IMF Working Paper – WP/02/195, December. Fasano, U. and Q. Wang, (2002), ‘Fiscal Expenditure Policy and Non-Oil Economic Growth: Evidence from the GCC countries’, IMF Working Paper – WP/02/201, November. Frankel, J. and A. Rose (1998), ‘The Endogeneity of the Optimum Currency Area Criteria’ The Economic Journal, 108, 449, July. Gulf Research Center (2005), ‘GCC Integration – More Rhetorical than Real?’ www.gulfinthemedia.com June 27th.
24
Gulf Research Center (2005), ‘Co-operation Council for the Arab States of the Gulf (GCC Charter)’ www.gulfinthemedia.com Sept. 15th. Jadresic, E. (2002), ‘On a Common Currency for the GCC Countries’ IMF Policy Discussion Paper – PDP/02/12, December. Laabas, B. and I. Limam (2002), ‘Are GCC Countries Ready for Currency Union?’ Arab Planning Institute – Kuwait, April. Mongelli, F. P. (2002), ‘New views on the Optimum Currency Area Theory: What is EMU Telling Us?’ European Central Bank – Working Paper Series, 138, April. Okogu, B. (2003), ‘The Middle East and North Africa in a Changing Oil Market’ IMF Publications. Washington: International Monetary Fund. Sept. 5. . Popescu, M. and S. Mustafa (2001) ‘The Gulf Monetary Unification: Opportunities and Challenges’ Arab Bank Review, 3, 1, April. Rutledge, E. (2005), ‘Will Record Oil Prices Help Economic Diversification?’ www.gulfinthemedia.com July 28th. Rutledge, E. (2006), ‘The dollar declines, while the euro shines’ www.gulfinthemedia.com May 19th. Sturm, Michael and Siegfried Nikolaus (2005) ‘Regional Monetary Integration in the Member States of the Gulf Cooperation Council’(June 2005). ECB Occasional Paper No. 31, June. Available at SSRN: http://ssrn.com/abstract=752091 Tavlas, G. S. (1993), ‘The New Theory of Optimum Currency Areas’ The World Economy, 16, 6, 663-685. Tavlas, G. S. (1994), ‘Theory of Monetary Integration’ Open Economies Review, March 5, 2, 211-230. Tower, E. and T. Willett (1976), ‘The Theory of Optimum Currency Areas and Exchange Rate Flexibility’, Special Papers in International Economics, Princenton University, 11. Vaubel, R. (1976), ‘Real Exchange-Rate Changes in the European Community: The Empirical Evidence and Its Implications for European Currency Unification, Weltwirtschaftliches Archiv, 112, 3, 429-70. Willett, T. and C. Wihlborg (1999), ‘The Relevance of the Optimum Currency Area Approach for Exchange Rate Policies in Emerging Market 25
Economies’, in R. J. Sweeney, C. G. Wihlborg, and T. D. Willett (eds), Exchange-Rate Policies for Emerging Market Economies (Boulder, CO: Westview Press). Willett, T. (2003), ‘The OCA Approach to Exchange Rate Regime’, in D. Salvatore, J. Dean, and T.Willett (eds), The Dollarization Debate (Oxford: Oxford University Press). Willett, T. (2003),’Optimum Currency Area and political Economy Approaches to Exchange Rate Regimes: Towards a Framework for Integration’ Current Politics and Economics in Europe, 17, 1, 25-52. Willett, T., O. Permpoon, and C. Wihlborg (2007), ‘Endogenous OCA analysis and the Early Euro Experience’, Claremont working paper.
26
