Grenada
Ex Post Assessment of Longer-Term Program Engagement

This paper discusses Ex Post Assessment of Longer-Term Program Engagement on Grenada. Grenada’s engagement with the IMF played an important role in supporting the small island economy after it was buffeted by major adverse shocks. IMF support catalyzed substantial donor aid in the wake of unprecedented damage from two hurricanes and provided additional resources when the global crisis hit. Key reforms were also advanced, including the implementation of a value-added tax and strengthening of the nonbank regulatory framework. Executive Directors have agreed that a new program should support urgently needed fiscal consolidation, promote faster and more inclusive growth, and focus on a few macro-critical reforms.

Abstract

This paper discusses Ex Post Assessment of Longer-Term Program Engagement on Grenada. Grenada’s engagement with the IMF played an important role in supporting the small island economy after it was buffeted by major adverse shocks. IMF support catalyzed substantial donor aid in the wake of unprecedented damage from two hurricanes and provided additional resources when the global crisis hit. Key reforms were also advanced, including the implementation of a value-added tax and strengthening of the nonbank regulatory framework. Executive Directors have agreed that a new program should support urgently needed fiscal consolidation, promote faster and more inclusive growth, and focus on a few macro-critical reforms.

Introduction

1. Since 2006, Grenada has had extensive engagement with the Fund (Table 1). It is the first member of the Eastern Caribbean Currency Union (ECCU) to come under the Fund’s policy of longer-term program engagement which came into effect in early 2003.1,2 The period of engagement includes an initial PRGF arrangement that was approved in April 2006, which was augmented twice and extended before it concluded in April 2010; and a follow-on ECF arrangement that was approved in April 2010 and expired in April 2013. Grenada also made emergency purchases from the Fund, in the aftermath of several severe tropical storms/hurricanes in 2003/2004. Before 2006, the last arrangement was under the Extended Fund Facility (EFF) and was cancelled in 1984.

Table 1.

Grenada: Arrangements with the Fund, 2006-20131/

(In thousands of SDRs)

article image
Source: Finance Department

Under the new architecture for low-income country lending, PRGF arrangements were converted into ECF arrangements on January 7, 2010.

2. The paper is organized as follows. It first reviews macroeconomic developments, including events leading up to the program and then under the two programs, and then discusses program design and performance. It concludes with a discussion on lessons learned and options for future engagement. The EPA finds that future IMF engagement with Grenada would be beneficial including by supporting the authorities’ efforts to achieve fiscal sustainability and higher growth.

Macroeconomic Performance, Program Design, and Assessment

A. Background

3. Grenada experienced relatively strong growth and improvements in living standards for several decades following independence (Box 1). Growth averaged 4.5 percent over 1980-1999, exceeding the world average by over 1 percentage point, supported by preferential access to EU markets for its agricultural products and Grenada’s initial entry into the tourism industry. Per capita income increased steadily allowing Grenada to move into the upper-middle-income group.3 Its membership in the ECCU has provided an important anchor for price and exchange rate stability and financial development. With a population of only about 105,500, Grenada is considered to be a “micro state” and remains PRGT-eligible due its vulnerabilities.4

4. Since 2000, economic growth slowed significantly and became more volatile (text figure, Box 2). Expansionary fiscal policy, to counteract slowing growth, led to a rapid buildup of debt by 2002. While the additional spending was mainly on capital projects, the impact on growth was limited reflecting that the multiplier is estimated to be relatively low.5 In addition, the rise in indebtedness severely constricted fiscal space to counteract shocks and put a further drag on growth. Moreover, the country which had previously been thought to be outside the hurricane belt was hit by hurricanes Ivan and Emily in 2004 and 2005, respectively which wreaked unprecedented damage.6

A01ufig01

Economic Growth in the Past and Present

(in percent)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Sources: ECCB and Fund staff estimates.

Grenada’s Economy and History at a Glance

Grenada is a largely tourism-based, very small, open economy that is highly vulnerable to external shocks. Over the past three decades, the economy has shifted from the once dominant agricultural sector into services, which now account for over 70 percent of GDP. These include several fast growing sectors: education services, through an offshore medical school (St. George), and communications. Tourism and education are the main contributors to GDP but in comparison to other ECCU countries, Grenada appears to have fallen behind in past decades (see Figure 2).

A01bx01ufig01

Structure of the Economy

(in percent of GDP)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Grenada is a member of the Eastern Caribbean Currency Union (ECCU). The ECCU is one of four regional currency unions in the world.1 In comparison to the other islands in the union, Grenada is in the middle range in terms of population and economic size. Apart from Grenada, the union comprises five other independent countries (Antigua and Barbuda, Dominica, St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines), as well as Anguilla and Montserrat, which are overseas territories of the United Kingdom. The union shares a common currency which is pegged to the U.S. dollar since 1976, has provided a strong anchor for macroeconomic stability and low rates of inflation, and has facilitated financial system development.

Grenada gained its independence from the United Kingdom in 1974. The decade after independence was marked by civil unrest that culminated in the violent overthrow of the government. This triggered a U.S.-led invasion in 1983 after which a constitutional government was restored. The country has since enjoyed political stability, with a strong democratic tradition rooted in labor unions.

A01bx01ufig02

ECCU - 6: GDP, Population and Real GDP per Capita

(2012; bubble size is proportional to GDP per capita in U.S. Dollars)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Sources: ECCB and Fund staff estimates.
1 In addition to the European Economic and Monetary Union, there are two regional currency unions in Africa: the Central African Economic and Monetary Community and the West African Economic and Monetary Union.
Figure 1.
Figure 1.

Grenada: Key Indicators

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Figure 2.
Figure 2.

Grenada: Program Performance—Selected Economic Indicators

(Projections and Outcomes, 2006-11)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

1 Due to a major revision in the GDP data, the program projections have been rescaled by a constant factor.

Growth Challenges

Small states are subject to a number of macroeconomic constraints which pose challenges to growth.1 Challenges are particularly marked for “micro states,” like Grenada. These include burdens linked to a combination of indivisible fixed costs and diseconomies of scale. In the public sector, this results in higher costs (including a larger share of employment) and limited institutional capacity; in the private sector, in concentrated market structure and a lack of diversification; and in trade, in high transport costs. It also affects exposure to natural disasters, which as recently evidenced in Grenada can be very large in relation to the economy.

Grenada’s growth slowed markedly since 2000 and volatility increased. The slowdown was triggered by several factors that also affected other ECCU members, including an erosion of trade preferences with Europe, terms of trade shocks; reduced fiscal space; emigration of skilled labor. Increased natural disasters also contributed to lower growth and increased fiscal costs. While the tourism sector has been an important engine of growth, the concentration in this sector also brought with it increased vulnerability due to the heavy reliance on imports and susceptibility to global shocks, particularly a decline in tourist arrivals from advanced countries, mainly the U.S. and U.K., and via higher fuel costs (which can affect travel).2 It is also fiercely competitive, vying with neighboring Caribbean countries as well as more distant destinations.

The growth slowdown unmasked deep-rooted competitiveness problems. These problems have translated into high current account deficits, large indebtedness vis-à-vis the rest of the world, and more generally an unsustainable external position, in part due to weak public finances. Other indicators point to an erosion of competitiveness including a decline in tourism market share and a decline in total factor productivity. Cost disadvantages arise from, high labor costs, in part due to the high degree of unionization, high severance costs, and electricity costs, which are also among the highest in the world. Trade protectionism is also relatively high, as is the cost of credit.

The private sector remains underdeveloped, which limits growth and job creation. Lack of access to credit is one issue, including high interest rates; as well as non-financial constraints including high electricity costs and tax rates, an inadequately educated work force, exacerbated by high emigration rates, and trade regulations. Given the need to reduce the role of the public sector, addressing these constraints will be key.

1 See Macroeconomic Issues in Small States and Implications for Fund Engagementhttp://www.imf.org/external/np/pp/eng/2013/022013.pdf.2 See discussion in The Eastern Caribbean Economic and Currency Union, IMF, 2013 and The Caribbean, From Vulnerability to Sustained Growth, IMF, 2006.

5. In the wake of the devastation of two hurricanes, Grenada engaged with the Fund in 2006.7 The program was successful in catalyzing support from the World Bank and the Caribbean Development Bank (CDB), and in quickly marshalling substantial donor financing (25 percent of 2004 GDP) to help rebuild the country after the hurricanes. It also facilitated a cooperative debt restructuring with creditors (Box 3).

B. Macroeconomic Performance under the Programs8

6. The objectives of the programs were broadly the same, with the overarching aim of achieving debt sustainability. The initial program agreed in 2006 aimed to restore fiscal and debt sustainability, promote high economic growth, reduce vulnerabilities (related to extreme weather events and weaknesses in the financial sector), and alleviate poverty. The process of reconstruction posed an initial challenge to program implementation, which was more profoundly challenged by the global financial crisis. The successor, approved in 2010, aimed to continue the reform program (ensuring fiscal and debt sustainability, reducing vulnerabilities in the financial sector and generating high and sustainable growth to encourage private sector-led growth and reduce poverty) and mitigate the impact of the global crisis. While it had initial success in improving macroeconomic management, the program was put on hold in mid-2011.

7. The global financial crisis took a heavy toll on Grenada’s economy. Growth declined by some 4 percent while Grenada was under the two Fund-supported programs. While reconstruction after the hurricanes initially led to a rebound in the economy, the global recession brought on a severe contraction in growth. This outcome contrasted to initial medium-term projections of an increase of 4 percent and was also well below growth of other ECCU countries and other Caribbean islands, as tourist arrivals declined even more than in other ECCU countries (Figure 1).9 Unemployment and poverty rose sharply. Based on the labor force survey released in September 2013, unemployment was estimated to be 33.5 percent (with youth unemployment as high as 55.6 percent) and continues to be one of the highest in the region.

8. During the “assessment period,”8 the authorities were able to reduce the primary fiscal deficit, but remained at risk of a debt crisis (Figure 1). The arrangement approved in 2006 called for a substantial upfront adjustment in the primary fiscal balance, requiring an underlying adjustment of about 4½ percentage points of GDP, and a decline in the debt ratio to 60 percent by 2015 based on long term growth of 4 percent. The successor assumed a substantial reduction in the primary deficit (from 3.8 percent in 2009 to close to balance and then a surplus in 2011) and a further ratcheting down of the debt ratio to 60 percent by 2020, with an assumption of 4 percent for longer term growth. The outturn under the two programs was rather different (see text table). In 2006-2007 there was a widening of the primary deficit reflecting the time needed for the reconstruction from the hurricanes, followed by a narrowing from 4.7 percent in 2007 to 1.0 percent in 2010, although the continued primary deficits and a sharp reduction in growth contributed to a further rise in the debt ratio during this period.10 Expansionary fiscal policy in 2011, when the program went off track, then led to a rapid fiscal deterioration, reflecting higher expenditures, as well as a decline in the revenue to GDP ratio. To finance the deficits, the government relied increasingly on the issuance of short-term government paper.11 Domestic and external arrears also increased.12 Total public sector debt skyrocketed to 108 percent by end-2012, substantially above the average for the ECCU region (85 percent). Faced with a severe fiscal crisis and liquidity shortages, the new government13 announced in March 2013 that it would seek a “comprehensive and collaborative” debt restructuring.

Key Program Projections and Outcomes

article image

Refers to 4th review objectives since both 5th review and program approval of 2010 ECF took place at the same time.

The debt ratio outcome reflects a substantial revision in nominal GDP in 2010.

Public Debt

Grenada ranks among the most highly indebted countries, with a debt ratio over 100 percent of GDP. Most of the debt is to external creditors (about half of which is loans from the Caribbean Development Bank and restructured commercial debt). Treasury bills and restructured bonds account for the bulk of domestic debt.

Debt restructuring

Grenada’s debt more than doubled from about 35% at the end of the 1990s to about 80% of GDP by 2002 as growth slowed and the government borrowed heavily on capital markets to finance capital expenditures. The authorities made efforts to reign in debt; however, these efforts were subsequently derailed after the devastating effects of Hurricane Ivan rendered debt unsustainable, by both eliciting higher spending and slowing economic growth.

In November 2005, Grenada undertook a successful debt exchange. Participation reached over 90% of eligible debt (about US$237 million or about 40% of total public debt at the time). There was no reduction of principal and past-due interest was fully capitalized. The new bonds had a 20 year maturity and interest rates of 1% for the first three years, which increased thereafter. The lower interest rates implied a haircut in NPV terms of some 35%, and a reduction in debt servicing of about 83% during 2005-2008. However, the rise in the interest rate now poses a significant burden, particularly in light of low growth.1

In May 2006, the Paris Club agreed to reduce Grenada’s debt service to its creditors by more than 90% for the duration of the PRGF arrangement. Subsequently, in 2010, the authorities planned to request a stock treatment of their Paris Club debt and to ask non-Paris Club official bilateral creditors for similar treatment. However, the debt restructuring did not take place.

Factors contributing to high debt

A reduction in the primary deficit, together with higher-than-expected GDP growth, helped to reduce the public debt-to-GDP ratio by end-2008 (see figure). However, the collapse in growth, high primary deficits, and rising interest costs contributed to an increase in debt ratios thereafter. Total public sector debt skyrocketed to 108 percent by end-2012. Faced with a severe fiscal crisis and liquidity shortages, the new government2 announced in March 2013 that it would seek a “comprehensive and collaborative” debt restructuring.

A factor that sustained high debt levels until recently has been captive financial markets, including regional banks and social security schemes that provided loans and were the main buyers of short-term government paper. Banks exposure to the government accounted for about 3.3% of GDP (including holdings of government treasury bills, government bonds, and loans to the government) and national social security’s exposure accounted for about 8.8% of GDP at end-2012. The establishment of Petrocaribe and continued assistance from multilaterals (at least until 2010) also were sources of government financing.

A01bx03ufig01

Change in Central Government (CG) debt to GDP ratio

(contribution to change in percent, yoy)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

1/ Includes arrears, privatization proceeds and other discrepancies due to valuation changes.
A01bx03ufig02

Grenada’s Public Debt

(in million EC $)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Sources: Grenada authorities and Fund staff estimates.
1 The interest rate increased from 2.5 percent in 2011 to 4.5 percent in 2012–2013, and will then rise to 6 percent in 2014–2015, 8 percent in 2016–2017, 8.5 percent in 2018, and 9 percent in 2019–2025.2 Grenada’s general elections on February 19, 2013, resulted in a change of government. The incumbent National Democratic Congress (NDC) Party was defeated by the opposition New National Party (NNP) under the leadership of Dr. Keith Mitchell, who had governed the country during 1995–2008. The NNP won all 15 seats in parliament.

9. Inflation rates fluctuated widely, reflecting Grenada’s highly open economy. After the two hurricanes, inflation was high, especially fuel price increases. It then decreased sharply in 2009, as the global recession led to lower commodity prices; the introduction of the VAT in 2010 together with rising fuel costs elevated prices again. Since then, inflation and broad money growth steadily declined to 0.7 percent and 1.5 percent in June 2013 (yoy), respectively.

10. Grenada’s current account deficit remained high, raising concerns about sustainability and competitiveness. After reaching a decade high in 2005 amid strong reconstruction activity, the current account deficit declined somewhat afterwards but continued to stay at a high level in the past six years (23.8 percent on average during 2007-2012). While the deficit was mainly financed by offical transfers, government borrowing, and FDI, more recently, it was mainly financed through debt-related flows, including increases in banks’ foreign liabilities and external arrears. In addition, it did not appear that private domestic consumption had adjusted sufficiently downwards to the signficant terms of trade shock the country experienced since 2003. This has raised concerns about medium term sustainability and the need to improve competitiveness and the business climate (see Box 2).

A01ufig02

Imputed Reserves

(months of imports; CA balance, imputed reserves - %GDP)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: ECCB and Fund staff estimates.

11. The financial sector regulatory framework was strengthened, but vulnerabilities remain (Table 2, Box 4). Grenada was the first ECCU member to establish a single regulatory agency—Grenada Authority for the Regulation of Financial Institutions (GARFIN). The supervision of insurance companies was also strengthened through the new Insurance Act introduced in 2010. However, the collapse of the CL Financial Group in 2009 and the contraction in growth highlighted vulnerabilities in the sector.14 NPL’s increased and there was a decline in profitability. While, all banks had written off their exposure to the CL Group by end-2012, the low level of other provisions and further decline in profitability suggest the need for close monitoring of banks.15

Table 2.

Grenada: Financial Soundness Indicators, 2004–2013

article image

12. Some improvements were made in living standards while poverty continued to rise Table 3). The standard of living improved relative to 2003/2004 in terms of increased life expectancy, lower child mortality and a decrease in illiteracy, reflecting higher GDP per capita. Although income increased, Grenada lost ground in relation to GDP relative to other ECCU members due to16 lower growth rates in the past years. The incidence of poverty also rose, with the poverty headcount index rising by 5.7 percentage points to 37.7 percent in 2008 (relative to 2004),—by far the highest rate in the region.

Table 3.

Eastern Caribbean Currency Union: Social and Demographic Indicators

article image
Sources: World Bank, WDI database; ECCU Countries and IMF staff estimates. Historic values are taken from “Ratna, Robinson, and Cashin, “The Caribbean. From Vulnerability Sustained Growth”.

Shading in red indicates a deterioration; shading in green indicates inprovement.

Grenada’s Financial Sector

The banking sector in the ECCU countries is supervised by the ECCB. Licensing of banks is the responsibility of the each country’s Ministries of Finance. The non-banking sector is under the purview of the each country’s supervisory authorities.

Banking sector

Currently, 5 banks are operating in Grenada, of which 4 are foreign owned. Bank assets are around 133 percent of GDP. As indicated in Table 2, banks have been holding above the regulatory capital requirement (on average CAR of 13.1 percent versus the required 8 percent) and sufficient liquid assets (cash, balances held at the central bank and treasury bills) in relation to current regulations (on average 28.5 percent versus the required 20 percent).1

Non-banking sector

In the nonbanking sector, Grenada’s regulatory agency, GARFIN, currently supervises 10 credit unions, which hold assets equal to 18.3 percent of GDP, 24 insurance companies holding assets worth 11.8 percent of GDP, 40 pension funds, 2 money service providers and 1 Building and Loan Association (B&L) which is under judicial management due to its exposure to the CL Financial Group. There are no off-shore banks in Grenada.

1 Higher capital ratios need to be interpreted with caution, however, particularly when juxtaposed against loan classification and provisioning requirements that are less stringent that international standards.

C. Program Design and Assessment

13. The track record of completing reviews under the two Fund-supported programs was uneven. As shown in the timeline (Box 5), the initial PRGF arrangement17 approved in April 2006 was extended by one year reflecting a nearly two-year delay in completing the first review. A contributing factor in the delay was the challenge posed by the reconstruction from Hurricanes Ivan and Emily, which took longer than originally expected and which resulted in fiscal slippages. After the first review was completed in 2008, the subsequent reviews were completed on a relatively timely basis, although most required waivers, reflecting in part the challenge posed by the global recession. The successor ECF arrangement was approved in April 2010 but only the first review was completed. While the program was on track to meet (or nearly meet) the targets for the second review, it was put on hold in mid-2011 when the authorities informed staff of their intention to pursue a debt restructuring. Subsequently, they undertook expansionary fiscal policy in the face of weak growth and amid a difficult political situation in the run up to elections (see further elaboration below). The program expired in April 2013.

Timeline of the ECF-Supported Programs

On April 17, 2006, the Executive Board approved a three-year PRGF arrangement. The objectives were to restore fiscal and debt sustainability, promote high economic growth, reduce vulnerabilities and alleviate poverty.

The program went off track quickly. Policy slippages, which reflected in part the additional time needed for reconstruction in the aftermath of Hurricanes Ivan and Emily, capacity constraints, and issues with an unlicensed indigenous bank were key reasons. The first review was not completed until July 2008. The program was extended by one year, augmented and rephased. The remainder of the program was completed broadly within the expected time frame.

During the course of the program the country was buffeted by a number of shocks. These included the commodity price shock and the global financial crisis that followed later in 2008. The Fund-supported program responded flexibly through the use of waivers, adjustments to performance criteria, and augmentations.

A successor ECF arrangement was approved on April 18, 2010, but went off track in mid-2011. The program objectives remained broadly unchanged. The first review was completed on time. However, the program was put on hold before the second review was completed. Subsequently, the authorities indicated that they wished to pursue expansionary policies. The arrangement was allowed to expire in April 2013.

A01ufig03

Original Program Schedule

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

14. Both Fund-supported programs were primarily focused in the fiscal area, reflecting the overarching aim of bringing debt down to a more prudent level. In the quantitative area, the main target was on the primary fiscal balance. In the structural area, fiscal reforms were mainly focused on public financial management. Reforms were also included to meet the objectives of reinvigorating economic growth primarily by promoting private sector activity; reducing vulnerabilities to natural disasters and safeguarding the soundness of the financial sector; and advancing social development by improving the social safety net (eventually this was combined with growth). However, there were only a few structural reforms in these areas. While the focus on the fiscal area for both quantitative criteria and structural measures was probably appropriate given unsustainable debt and large fiscal deficits; as discussed further below, more emphasis should have been given to growth-inducing measures.

15. The Fund-supported programs included a large number of structural reforms. While quantitative criteria were fairly standard in the 6-7 range, there were some 31 structural measures, primarily benchmarks, that were established as conditions under the 2006-10 and 2010-2013 arrangements.18 This translated into an average of approximately five structural measures per review (a total of six reviews were completed). While quantitative criteria remained broadly constant, structural conditionality was moderately front-loaded, with 15 structural conditions, including two Prior Actions, assessed during the first two reviews. However, overall performance in implementing these conditions was weak, with less than one-third of measures met on schedule, and nearly 40 percent never being achieved.

D. Quantitative Performance

16. Quantitative conditionality was broadly similar under both programs with the primary deficit, as the most important performance criterion. A floor on the primary balance was placed on the central government primary balance, excluding grants.19 A ceiling was also placed on the stock of domestic arrears.20 There was a standard criterion on zero accumulation of external payment arrears and on the stock of short-term external debt, and a ceiling on contracting and guaranteeing non-concessional external debt by the government. An additional criterion on bilateral concessional borrowing was added later to discourage the contracting of a large loan from China’s Export-Import Bank that posed a risk to debt sustainability. An indicative target on net bank credit to the government was also included.

17. Quantitative performance was uneven (Table 4). The primary deficit performance criterion was only met in about one-half of the reviews.21 In most cases the deviations from this criterion were substantial. The other quantitative targets were broadly met. The performance criterion on the stock of domestic arrears was met throughout the program. The performance criterion on external arrears was also met for the most part. The zero ceiling on short term debt, non-concessional debt, and bilateral concessional external debt was met at all times. The indicative target on net credit to the government was met with the exception of the first review.

Table 4.

Grenada—Quantitative Performance under the IMF Programs

article image
Source: Monitoring and Funds Arrangement (MONA) database and Staff Reports.

The first review under the 2006 program was delayed because of the time needed to address an unregulated bank, fiscal slippages, and the slow pace of structural reforms. The review was completed in June 2008.

Cumulative.

Adjusters were included to accommodate grants and other higher-than-programmed external disbursements, including on existing donor-financed projects and additional concessional financing.

Excludes PetroCaribe.

End-December 2008 figure for central government primary balance was corrected from -107.5 at the time of the 3rd review so that target was actually missed. Given the small size of the deviation the case was treated as de minimis misreporting.

The end-June 2009 target for the central government primary balance was relaxed by 1.4 percent of GDP (original program target was EC$28 mil).

End-November PC on domestic arrears was relaxed by EC$12.5 million (0.7 percent of GDP) due to tight liquidity situation and likelihood of several large disbursements would not take place until December.

18. A series of shocks contributed to target misses. The initial performance criterion on the primary balance was missed by a wide margin, reflecting in part higher-than-expected costs of reconstruction (Table 4). Revenues were also weaker than projected mainly due to lack of progress in structural reforms, as discussed further below (Figure 2). While the economy rebounded in 2007-2008, the June 2008 performance criterion on the primary balance was also missed on account of measures to alleviate the external commodity price shocks, high spending in the run-up to the elections in July 2008, and unbudgeted retroactive wage payments. While economic growth subsequently slowed in 2009 in response to the global slowdown, spending retrenchment helped meet the targets for the third review, however the fiscal target for the fourth review had to be revised as growth contracted more than anticipated. Major fiscal slippages recurred in the final review of the 2006-2010 Fund-supported program, due to expenditure overruns, with the primary deficit performance criterion missed by a wide margin. The performance criterion on the non-accumulation of external arrears by the central government was breached in several cases due to the late arrival of budgetary support and other procedural weaknesses. These arrears were cleared and the authorities undertook measures to strengthen debt management, with MCM technical assistance.

19. Staff requested waivers on fiscal targets mainly on the basis of remedial measures to be taken by the authorities to reduce spending. The remedial measures included efforts to strengthen expenditure control, particularly capital spending, including by limiting the use of special warrants and the issuance of circulars which were prior actions (in several cases). In approving the waivers, the Board recognized the global crisis had a significant adverse impact on Grenada and regretted the slippages and called for strong actions to improve expenditure control. As discussed below, the authorities were able to make progress in reducing the deficit while the programs were on track.

20. Under the 2010-2013 successor arrangement, the authorities intended to undertake a more aggressive fiscal adjustment to place the public debt-to-GDP ratio firmly on a downward trajectory. All the quantitative targets were met for the first review. However, the composition of spending was not as envisaged, as planned savings on goods and services did not materialize (partly reflecting higher-than-expected fuel prices) and were offset by containing transfers and subsidies and limiting capital outlays. The second review was not completed, although it appears that the most of the performance criteria could have been met (although the primary surplus and domestic arrears target may have been missed by a small margin). The second review was postponed because the authorities indicated to staff that they wished to pursue a debt restructuring. Subsequently, the program could not be brought back on track because the authorities decided to undertake fiscal stimulus measures in the face of continued weak growth and a complicated political situation (thin parliamentary majority) ahead of elections.

21. The primary fiscal balance improved but remained in deficit. While it is difficult to assess the adjustment that took place given the shocks to the economy, staff’s estimates of the structural primary deficit suggest that there was a reduction in the underlying primary deficit on the order of 2 percentage points between 2006 and 2012, with a larger adjustment taking place when the programs were on track (2008-2010) (see Box 6). The fiscal impulse analysis suggests that the fiscal stance was contractionary during most of the period during which the two Fund-supported programs were in place although it became expansionary in 2011 after the successor program was put on hold. The impact on growth of the adjustment was probably limited, however, since the fiscal multiplier is low, reflecting the high propensity to import and the high level of indebtedness (see Box 6).

A01ufig04

Figure. Changes in General Government Primary Expenditure

(Percent of GDP)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: IMF Staff Calculations

Fiscal Adjustment

A sizeable fiscal adjustment was needed due to the large debt overhang. The high debt-to-GDP ratio in the context of the regional currency arrangement, constrained the scope for countercyclical policy response. A flatter adjustment path would have required more external financing. Despite augmentation of Fund resources, front-loading of disbursements and the catalytic role of Fund financing, Grenada faced a tight resource envelope which ultimately required an adjustment.

bx06ufig01

Grenada: Fiscal balances

(In percent of GDP, unless indicated otherwise)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: Country authorities and staff estimates.1/ A positive (negative) fiscal impulse implies expansion (contraction).

The shocks that buffeted Grenada make it difficult to discern precisely the extent of the adjustment. The structural primary balance may provide better insight into the underlying fiscal position, since this takes into account cyclical factors and potential GDP.1 Staff estimates suggest that there were significant structural primary deficits during the period, which declined from 4.5 percent in 2006 to around 2.7 percent in 2012. The fiscal impulse analysis suggests that fiscal policy was contractionary during most of the time the program was active, but then became expansionary in 2011.

Grenada’s fiscal multipliers are small suggesting that a contractionary stance had a limited adverse impact on growth. Recent research on Caribbean islands and ECCU by Guy and Belgrave (2012), Gonzalez-Garcia et al. (2013) and Acevedo et al. (2013), respectively find that fiscal multipliers are low and attribute this to the high debt and high degree of trade openness.2 Gonzalez-Garcia et al. (2013) finds that the long-run multipliers of taxes and consumption are statistically insignificant while public investment has a long run multiplier of 0.6. This suggests that while the policy should center on long-term solvency rather than short-term policies given the limited benefits, if some countercyclical policies are implemented they should focus on capital spending.

A second wave of debt restructuring might have allowed for a more gradual adjustment by tempering the debt overhang, if it involved a principal reduction. The commercial debt exchange, which was completed in November 2005, did not involve any write down of the principal and the past-due interest was fully capitalized (See Box 3). Earlier recognition of the need for a hair cut may have allowed for a more gradual fiscal adjustment under the program. A recent IMF paper finds that debt restructurings have often been too little and too late, thus failing to restore debt sustainability and market access in a durable way.3

1 It is important to note the difficulties associated with estimating potential GDP and elasticities of revenue and expenditure. The estimates follow FAD’s methodology (see Annex 1).2 J. Gonzalez-Garcia, A. Lemus, and M. Mirkaic, “Fiscal Multipliers in the ECCU” IMF Working Paper 13/117; and K. Guy and A. Belgarave, “Fiscal Multipliers in Microstates: Evidence From the Caribbean,” International Advances in Economic Research, February 2012.3 IMF Board paper “Sovereign Debt Restructuring—Recent Developments and Implications for the Fund’s Legal and Policy Framework” http://www.imf.org/external/np/pp/eng/2013/042613.pdf.

22. Reductions in primary spending came mainly through cuts in capital expenditures. There were sizeable cuts in capital spending during the period 2006-2012 (in part reflecting the expected decrease in capital spending after the recovery from the hurricanes).22 The government faced difficulties in containing the wage bill as envisaged under the program and current spending rose significantly during the 2006-2012. While reducing spending was important for adjustment, to preserve the prospects for long-term growth and job creation, cuts in current spending rather than deeper cuts in capital spending would have been preferable.23 Social spending, including pensions, increased in relation to GDP.24 However, revenues, which were expected to assist the adjustment, remained relatively flat in relation to GDP.

A01ufig05

Government social spending, in percent of GDP

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

23. Little progress was made toward the program objective of reducing debt vulnerabilities. At the time the program was approved in 2006 the aim was to bring debt to 60 percent of GDP (in line with the ECCB’s recommendation) by 2015, which assumed strong growth on the order of 4 percent. Although the public debt-to-GDP trajectory declined as the economy recovered, this trend was reversed in 2009 in the wake of the global financial crisis. The growth outturn was much worse than expected and debt sustainability remained out of reach in the absence of a very large fiscal adjustment.25 As it was, the continuation of fiscal primary deficits contributed to the worsening outturn.

24. A possible large loan from China’s Export-Import Bank posed a major risk to debt sustainability through most of the assessment period. Estimates ranged from about US$85 million to about US$100 million or about 11-13 percent of 2011 GDP. In view of staff’s concern that such a large loan would put the objective of restoring debt sustainability at risk, particularly if the expected growth impacts did not materialize, the PRGF-supported program targets were revised during the First Review in 2008 to include a ceiling on bilateral concessional external debt. It was noted at the time of the first review in 2008 that the ceiling would be revised upward later should the authorities proceed with a loan that does not jeopardize debt sustainability.26

Factors affecting performance

25. Several factors contributed to the weak quantitative performance. In particular, the global recession made fiscal adjustment extremely difficult in the face of declining growth. The program responded flexibly through several augmentations, rephasing of disbursements, modifications to performance criteria, as well as program adjustors to accommodate external financing.27 However, there were clearly difficulties in setting targets in an economy which is highly vulnerable to shocks and high growth volatility. In general, growth projections were often too optimistic under the 2006-10 Fund-supported program (see Figure 2) and longer term growth projections may have been too high at 4 percent.

26. Capacity constraints and political pressures were also a factor. As discussed below, progress on reforms was constrained by capacity limitations. Also, since expenditure overruns were a major factor in missing the fiscal targets, more focus needed to be placed on expenditure control, especially the wage bill. As noted, the Fund-supported programs resorted to prior actions aimed at keeping capital expenditure below levels agreed upon in the budget through circulars. These approaches had a limited term impact and were vulnerable to being overridden by the MoF.28 Also, placing too much emphasis on cutting capital spending may have not been helpful for growth, since the fiscal multiplier on capital spending is estimated to be different from zero. The successor arrangement aimed to strengthen the spending framework through quarterly meetings of the line ministries and closer monitoring of quarterly indicative targets, although it also relied on the use of circulars to contain budget spending.29 It is also evident that political pressures were also a factor. Election-related pressures contributed to fiscal slippages at the start of the first program and in the successor arrangement moving off track in 2011. Also revenue efforts were undermined in part by the wide use of tax exemptions and concessions and weak tax administration.

E. Structural Performance

27. The structural reform agenda supported by Grenada’s two PRGF/ECF arrangements over 2006-2013 was ambitious. Some 31 structural measures, nearly all taking the form of structural benchmarks, were established as conditions under the 2006-2010 and 2010-2013 arrangements and assessed through a total of six reviews—an average of approximately five measures per review.30 Structural conditionality was moderately front-loaded, with 15 structural conditions, including two Prior Actions, assessed during the first two reviews. However, overall performance in implementing these conditions was weak, with less than one-third of measures met on schedule, and nearly 40 percent never being achieved.

28. Structural conditionality focused almost exclusively on public financial management, with more than three-quarters of the measures being established in this area. This reflected the 2006-2010 program’s aim, at its outset, to “sharply reduce public debt from its present inordinately high level (128 percent of GDP at end-2005)”.31 While the initial program strategy called for fiscal consolidation to be achieved through an even distribution of revenue and expenditure measures, structural conditionality in the fiscal area was in the event more heavily weighted toward revenue measures.

29. Some two-thirds of measures established in the two programs during the assessment period were in the revenue area. This included elimination of tax holidays and their replacement by a system of accelerated depreciation, and the introduction of a VAT. The focus on eliminating tax holidays reflected the strongly held view of the team that “[t]he current system of tax incentives has proved ineffective at promoting investment…, imposes a heavy burden on public finances, and is very cumbersome to administer.”32 Indeed, it was noted that foregone revenues through this practice amounted to more than 10 percent of GDP. There was also a noteworthy focus on improving customs administration. An additional possible explanation for the emphasis on formal program conditionality in the revenue was the recognition that action in this area would prove politically challenging. As we note below, if this was indeed a motivation, it proved prescient.

30. While an emphasis on fiscal measures was probably appropriate, more attention should have been devoted in practice to growth-oriented policies and efforts to reduce vulnerabilities. Only a few structural conditions dealt with the financial sector, the investment climate, or directly with poverty alleviation. The absence of conditions in these areas appears to contradict the strategy presented at the time of the approval of the arrangements, whereby “[t]he first year of the program targets measures to improve the investment climate, and the second and third years are expected to focus on removing policy-induced distortions in labor and financial markets. Program documents also refer to efforts toward mitigating vulnerabilities, in particular a number of initiatives to improve preparedness for extreme weather events as well as to address weaknesses in the financial sector.

Performance under the Fund-supported programs

31. Performance in implementing structural reforms during the 2006-2010 program was weak (see Appendix I). Of a total of 26 measures, only 7 were met on schedule, 9 were implemented with delay, and 10 were not met or dropped. Progress was particularly weak with respect to revenue measures, and especially dismal in an area identified as key to improving fiscal performance—limiting tax incentives for investors and replacing them with a more neutral and transparent system based on accelerated depreciation. Here, measures were established at several stages in the process of removing tax holidays, with conditions repeatedly being missed and reset.33 As a result of the heavy (and unsuccessful) focus in this area, only 20 percent of revenue measures were met as scheduled, and 5 out of the 7 measures that were never met were in the revenue area. In contrast, expenditure measures were about twice as likely to be met over the lives of the two programs being assessed.

32. Many of the instances in which structural measures were implemented (benchmarks met or implemented with delay) appear to have been cases where the criteria for success were weak. These include cases in which the relevant structural benchmark called for a process to be “initiated”, legislation submitted to Parliament, or involving a matter of internal administration within the public service. In contrast, cases in which the measure was focused on a major final outcome (e.g., establishing a public procurement authority or repealing tax incentives) were typically not successful.

33. The implementation of a VAT in 2010 represented a major accomplishment in the program of structural fiscal measures. The introduction of a VAT is both administratively complex and politically challenging, particularly in the Grenadian context in which a previous attempt to establish a VAT had failed. Owing to its efficiency advantages and potential to raise significant amounts of revenue, the successful effort to put in place a VAT represented a major and durable success of the 2006-2010 program. It has been noted, however, that the VAT’s yield has been eroded by the practice of granting ever-expanding exemptions. Halting and reversing this process should thus be a priority of future Fund programs.

34. Grenada also made important progress on structural reforms in areas that contributed to reducing vulnerabilities, but which were not the subjects of formal program conditionality. This included the introduction in 2006 of a more broad-based income tax—the National Reconstruction Levy (NRL)—and the establishment of the Grenada Authority for the Regulation of Financial Institutions (GARFIN) to regulate the non-bank financial sector. Important steps were also taken to strengthen the building code and improve its enforcement and Grenada also signed up for the World Bank’s Caribbean Catastrophe Risk Insurance Facility (CCRIF).

35. Debt management represented an area which staff reports cited as a source of vulnerability, yet was largely absent from program structural conditionality. Grenada’s debt is both high and financed mainly through short-term instruments, increasing rollover risk. However, with the exception of a structural benchmark on the establishment of a debt management unit at the Ministry of Finance (met at the third review of the 2006-2010 PRGF/ECF arrangement), program conditionality made little attempt to address the situation. Indeed, the 2012 Article IV staff report stressed the continuing vulnerabilities resulting from weak debt management capacity.

36. Efforts to strengthen the investment climate appear to have been short-lived. The initial request for the 2006-10 arrangement noted that “[t]he paucity of new private investment—particularly in the tourism sector—is a major concern of the authorities”. Accordingly, the first year of the program included two structural measures in this area, one aiming at simplifying land use planning and the other involving development of an action plan to improve Grenada’s ranking in the World Bank’s Doing Business indicators. Neither structural benchmark was met when assessed at the time of the second review and they were then permanently dropped.34 Only one other measure related to the investment climate was promised over the remainder of the program: establishment of a single registrar of companies (met at the third review).

37. Implementation of structural conditionality in Grenada’s successor ECF-supported program, approved in 2010, made a strong start. All five structural benchmarks were met on schedule. While it could be argued that many of these measures were either initial steps in more complex processes, or otherwise less ambitious and politically non-controversial, performance under the first review of the 2010 arrangement appeared more promising than under the 2006-2010 arrangement.

38. Recent assessments of Grenada’s economy have noted that the lack of progress in structural reforms is taking a serious toll on competitiveness. In particular, the 2012 Article IV staff report noted that the environment faced by the tourism sector was becoming more challenging, and that rigid labor market institutions were contributing to a trend toward rising unit labor costs and falling profitability. In this respect, it is noteworthy that while the request for the 2006 PRGF/ECF arrangement noted that reform of labor markets would be an objective of the latter years of the program, it is not clear whether the staff had yet devised specific reforms that they intended to propose. It is thus impossible to evaluate whether this approach would have been feasible and effective in promoting growth.

39. The structural reform agenda of the ECF program was supported by a substantial commitment of technical assistance. Technical assistance, which had been almost negligible before the beginning of the ECF arrangement in 2006, increased sharply over the 2006-2010 period. Moreover, the composition of technical assistance corresponded closely to the emphasis in Fund conditionality on public financial management reforms, with nearly two-thirds of TA being in this area. Nearly 85 percent of fiscal TA was in the area of revenue administration. The introduction of a VAT in 2010 appears to have been supported especially strongly by IMF assistance in revenue administration.

A01ufig06

Evolution of IMF Technical Assistance, by Department

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

A01ufig07

Composition of IMF Technical Assistance, 2006-10

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

40. The IMF’s efforts to promote macroeconomic stability were complemented by an active program of engagement by other multilateral institutions, notably the World Bank and Caribbean Development Bank (CDB). The World Bank’s lending program in Grenada included components that aimed to strengthen public sector institutions, including those involved in revenue mobilization, and service delivery, as well as promote inclusive growth. This included programs in the last 4-5 years which provided technical assistance in targeted areas, including VAT/customs reform; public sector modernization, with a focus on streamlining employment in ministries; skills and inclusive growth, involving a technical assistance grant to provide skills training to improve prospects for employment. The Bank is supporting the agriculture sector of Grenada through a program of grants to small farmers to help them cope with the recent global financial and food crisis and weather fluctuations. While progress was made in these areas, capacity constraints were a factor in reform implementation. In addition, more extensive engagement by the World Bank, particularly in the financing of infrastructure investment, was constrained in recent years by a limited envelope for additional lending (US$20-30 million over 4 years), reflecting high levels of existing debt and Grenada’s status as a middle income country with limited access to IDA concessional resources. The CDB’s activities lent particular support to anti-poverty efforts as well as growth-enhancing projects focused on rural development.35 Recent projects have included a fast-disbursing Policy-Based Loan (PBL) aimed at providing budget support, as well as schools rehabilitation and reconstruction projects.

Factors affecting performance

41. The causes of the weak structural reform performance under the first Fund arrangement in the assessment period are difficult to identify with precision. Staff reports in this period strongly emphasize the “homegrown” nature of the program, suggesting that ownership was strong. It is possible, however, that engagement was uneven across the range of structural reforms, with highly controversial measures, such as the repeal of tax incentives, characterized by weaker ownership. (Indeed, the decision in 2010 to grant general VAT exemptions for the tourism and construction sectors suggested that the authorities’ retained a predilection for the use of such instruments.) At the same time, the authorities argued that the intense regional competition for investment meant that unilateral steps to repeal incentives would have a major adverse impact on Grenada’s ability to attract new investment.

42. In addition, the Grenadian authorities also were forced to respond to a series of unwelcome shocks emanating from the domestic and regional financial sectors. These included the failure of a previously unregulated bank, leading to the appointment of a receiver and attempts to liquidate the bank, leading in turn to protracted litigation. In addition, two regional financial institutions experienced severe financial difficulties. Although none of the institutions were considered systemically significant, the time consuming processes needed to respond to their failures is likely to have represented a significant distraction for members of Grenada’s small economic team, and thus impeded program implementation

43. It is also important to note the thin Parliamentary majority (one seat) of the government during the first two years of the program, which may have constrained unpopular measures. However, the subsequent landslide victory of the opposition party did not translate into a sea-change in program implementation.

Did reform agenda support program objectives

44. The poor record of implementation of structural fiscal measures is very likely a significant contributing factor in the failure of Grenada’s IMF programs to achieve a sustainable fiscal situation. While assessment is complicated by the exogenous shocks affecting the country, including the strongly adverse impact of the 2008 financial crisis, it appears that large structural deficits persisted over the assessment period, although there was some improvement (see above). In addition, most of the adjustment came through cutting capital spending, while the wage bill increased significantly, and revenue performance remained relatively flat.

45. There is also little concrete evidence to support a conclusion that Grenada’s investment climate has improved during the assessment period. Grenada’s level in the Doing Business Indicators’ Distance to Frontier index remained virtually flat over the 2006-2013 period, although in this respect its experience mirrors the other members of the ECCU.36 The lack of progress in improving the investment climate is unfortunate, given that attracting additional private sector investment—through means other than tax concessions—offers perhaps the only means for sustainably improving growth performance and moving away from a public-sector-led growth model that has clearly run its course.

Ranking in World Bank Doing Business Indicators Distance to Frontier Index: 2006-13

article image
Source: World Bank.

Excludes Anguilla and Monserrat.

Lessons and Challenges for Future Program Design

A. Lessons Learned

46. Grenada’s engagement with the Fund was important in supporting the country in the wake of unprecedented hurricane devastation and when it was hard hit by the global recession. The initial program helped catalyze essential donor financing from multilateral agencies and paved the way for a commercial debt exchange in 2005 and a Paris Club debt rescheduling in 2006.37 It also responded flexibly in the face of the shocks that buffeted Grenada, providing additional resources and allowing targets to adjust (although still requiring progress on deficit reduction given the large debt overhang). In addition, there was progress on important structural reforms. In the financial sector, supervision of the non-bank financial sector was strengthened with the establishment of a single regulatory unit—a first among ECCU members, although this accomplishment was not subject to formal conditionality in the Fund-supported programs. Furthermore, the authorities were able to implement the VAT which was a complex and difficult achievement. They also reformed the customs authority (under instruction of ASYCUDA World). The successor arrangement in 2010 had some initial success in improving fiscal management, but then it went off track.

47. In general, however, the Fund-supported programs were unable to deliver on key program objectives of high sustained growth and debt sustainability. The other aim of reducing vulnerabilities was also not met, although there was significant strengthening of the regulation of the non-financial sector. The series of shocks which buffeted Grenada during the period being assessed made achieving the programs’ goals much more difficult. While the government was able to achieve a significant reduction in the fiscal deficit during the assessment period, the continued deficits along with slowdown in growth led to a further rise in the debt ratio and Grenada continued to face debt distress under both programs. In addition, most of the adjustment came from cuts in capital expenditure, while structural components such as the wage bill expanded. Also, poverty remains widespread and unemployment is high.

48. Given the weak performance, the question arises whether the program design was realistic given political and institutional constraints. Although the authorities actively sought Fund and CARTAC technical assistance, as a micro state, they faced significant capacity constraints, which were even more heavily taxed by the adverse shocks that buffeted them.38 This limited their ability to address the ambitious reform agenda. In addition, the need to engage with social partners on reform measures necessitated more time for the authorities to implement measures (e.g., personal income tax reform). There was also probably too much emphasis on halting the practice of granting tax concessions, which was not productive given an apparent absence of ownership by the authorities in this area, who appear to have recognized that a unilateral abandonment by a single country would place it at a major competitive disadvantage. It might have been more realistic to have attempted to establish a framework for limiting incentives. In contrast, attempts to improve the investment climate were abandoned rather quickly, after modest initial setbacks, although measures in this area would have been important for growth.

B. Future Engagement

49. Grenada stands at a critical juncture and bold strategies to address the challenges are essential. High debt, low growth, and weak financial systems remain major problems that need to be tackled. No single bullet exists to address all of the challenges. What is clear is that past approaches of promoting growth through higher government spending have run their course.

50. A future program will need to take into account the specific characteristics of Grenada as a micro state—very limited capacity and high vulnerability to natural disasters and other exogenous shocks. To take account of capacity constraints, the program should focus on a few key macro-critical reforms with appropriate sequencing of technical assistance. A future program will also need to take into account the volatility of growth and exposure to shocks, including downside scenarios, and more realistic growth forecasts. Contingency planning should also feature in the budget in case a disaster hits. This could include identifying contingent measures to reduce spending if needed.39

51. A new arrangement will also need to establish clear ownership and a track record, given the weak past performance. This should involve setting comprehensive macro-critical prior actions and indicative targets on key fiscal variables in the period prior to Board consideration of the program.

52. Growth-related reforms need to be front-and-center. Given the growth challenges facing Grenada and the limitations to diversify as a small state, reforms should be targeted to business climate and regulatory reforms that are macro-critical, although it is recognized that these reforms take time to develop and bear fruit.40 The reforms can draw on recent staff work in this area and aim to increase competitiveness. Coordination with the World Bank and other development agencies would be essential. Ideally, a comprehensive growth strategy would be developed where public sector demand would be replaced with self financing private sector projects as the main engine of growth.41

53. Fiscal adjustment will be an unavoidable element given the extent of fiscal and external imbalances and the debt overhang, which puts a drag on growth. Adjustment will require a reduction in current fiscal expenditure, in particular the wage bill. On the revenue side, efforts should be made to halt further recourse to tax incentives, including addressing this at the regional level. Reforms should be aimed at supporting private sector job creation—building buffers to insulate against shocks and support macroeconomic stability; enabling the business environment (including appropriate tax and spending policies that support job creation); investment in infrastructure (transportation and energy) which can improve the business climate.”42

54. Debt restructuring would likely need to be part of Grenada’s strategy to return to sustainability, given the magnitude of the implied fiscal effort otherwise. The debt restructuring in 2005, while successful in terms of providing near term debt service relief through an extension of maturity and initial reduction in interest costs, turned out to be inadequate in the absence of a principal reduction. The continued primary fiscal deficits, along with the step up in interest rates placed increasing pressure on Grenada’s debt service and were major reasons behind Grenada’s inability to honor its debt, leading to another debt restructuring announcement in March 2013. The authorities should continue to engage in an early and collaborative dialogue with creditors so as to achieve a consensual restructuring that provides for medium-term debt sustainability alongside measures to strengthen debt management.

55. Regional collaboration can support program objectives and help overcome some of the constraints of small size, although it cannot be part of program conditionality. It may provide a more promising avenue for discouraging unsound measures, such as competitive tax exemptions. The Fund’s regional surveillance of the ECCU members provides a useful framework to help strengthen the supervisory and regulatory framework in the financial sector and to support further efforts toward fiscal coordination. In addition, CARTAC has successfully delivered regional technical assistance. Close collaboration with other IFIs (including the World Bank’s Comprehensive Debt Framework and regional catastrophe insurance) will also be important43 Given the exceptionally high costs of natural disasters, small states in the Caribbean should be seen as frontline candidates for support from climate-change funding, as global strategies for mitigation and adaptation to climate change become operational.

Annex I. The Views of the Authorities

The authorities welcomed the opportunity to reflect on the lessons of the Fund’s engagement in Grenada over the last decade.

Having reflected on the previous programmes, the key lessons are:

  • The programmes did not have sufficient focus on growth and job creation;

  • Cuts in capital expenditure especially investments in the productive sectors have hurt growth and job creation;

  • The protection of safety nets in periods of adjustment is essential;

  • Overly optimistic growth and revenue forecasts are not prudent and undermine the credibility of the programme; and

  • Overly aggressive primary balance targets are counter-productive as they are not only unrealistic but when they are not achieved render both the Country and the Fund in a negative light.

  • The capacity required for implementation was underestimated.

The authorities also offered the following comments:

The engagement with the IMF through two programmes was very helpful. The programmes supported recovery from two devastating hurricanes, severe financial sector dislocations associated with the failure of several financial institutions, and the impact of the global financial crisis. The 2006-2010 PRGF arrangement in particular had achieved its objectives of supporting fiscal adjustment and structural reform, and all the reviews were successfully completed. The establishment of GARFIN and introduction of the VAT during the programme period will have long-lasting positive impacts in improving financial sector supervision and enhancing revenue, respectively.

The 2008 financial crisis has had a major impact on the economy. It took a major toll on growth and represented a severe setback in their efforts to achieve long-term fiscal sustainability, including by eliciting renewed demands for tax concessions from hard-hit sectors of the economy. The sustained domestic slowdown stemming from these external developments was also a factor undermining revenue from the VAT introduced in 2010. Grenada’s ability to remain on track with its Fund-supported programme in this environment demonstrates the commitment of the authorities to their objectives.

Before entering into the successor arrangement, it would have been beneficial to reflect on lessons learned in the first programme. In retrospect, there was too much emphasis on reducing debt and the fiscal deficit and not enough on growth. A possible debt restructuring put the program on hold in 2011. Subsequently, serious political issues inhibited progress given the slim parliamentary majority, although a dialogue with the Fund was maintained.

The authorities concurred with the staff assessment that structural conditionality may have been unrealistic in terms of both what was politically feasible, and technically possible in the context of scarce implementation capacity. In this respect, it is important to emphasize the importance in future program design of focusing on a small number of “game changers”, rather a large number of worthwhile but non-critical objectives. In particular, the effort to limit tax incentives was misplaced, given the need for Grenada to compete with regional peers in a fiercely contested market for tourism-related FDI. While the substantial technical assistance provided by the Fund and others in the context of sectoral missions was welcomed, a longer-term technical advisor could have played a useful role by supplementing scarce domestic capacity.

A paramount lesson of the experience through two Fund-supported programs was the need to reinvigorate growth. A focus on achieving short-term budgetary objectives could prove counterproductive in fiscal terms, by undermining activity and thereby curtailing revenues, and would frustrate efforts to achieve the broader objectives of the program in fostering inclusive growth. It is especially important in this regard that efforts be made to create fiscal space for capital expenditure, aimed at addressing infrastructure bottlenecks and job creation, an important factor in addressing poverty. This could be assisted by efforts to raise revenues through efforts to strengthen tax administration and simplifying the tax system. A meaningful debt restructuring is essential to fiscal and debt sustainability and must not be taken lightly in the next programme.

Annex II. Estimating the Structural Primary Fiscal Balance in Grenada

This annex estimates the structural primary fiscal balance in Grenada, employing different methodologies (IMF and an augmented “hybrid” approach), and assesses the stance of fiscal policy with a particular focus on 2006-2012 (the program years).1 The results show that the structural primary deficit improved significantly during 2008-2010 when the program was on track but then deteriorated thereafter. In addition, the fiscal impulse analysis indicates that the expenditure impulse had a bigger contractionary impact on aggregate demand than revenues. It is recognized though that the estimation of the structural primary balance is subject to significant uncertainty particularly to the size of potential output.

A. Background and Motivation

Grenada has been subject to a continuous series of shocks since 2006 that impacted the fiscal balance. Reconstruction costs, in the aftermath of Hurricane Ivan and Emily, were much higher than expected under the program in 2006-2007, while the escalation of food and fuel price in 2007-2008 increased fiscal costs, and the global financial crisis and economic slowdown depressed FDI and tourism starting in 2008. While the reconstruction and other commodity price shocks were associated with a positive output gap, and accompanied by an “import boom” (i.e., imports in excess of “normal levels”), a substantial part of the import bill was born by the government through higher capital-related spending and food and fuel subsidies. In addition, these shocks were sometimes accompanied by a sharp decline in output growth but a positive output gap, for example growth decelerated sharply by 4 percent in 2006 in Grenada—after a very sharp reconstruction-related surge in 2005—resulting in a positive output gap. (Figure A1).

Figure A1.
Figure A1.

Grenada Primary balance, Output gap and Real GDP growth

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: Country authorities and staff estimates

Assessing the structural fiscal balance is particularly challenging in small islands, like Grenada, due to the difficulty of disentangling the shock and trend effects.2 To discern the fiscal effort, it is important to remove the automatic stabilizers and one-off factors, and focus on the underlying fiscal adjustment. Separating the transitory factors is complicated in small islands as they are subject to high frequency shocks and also may face a permanent decline in potential output. Ideally, the impact of one-off shocks such as natural disasters could be excluded but this is difficult in small islands which are adversely affected by a series of large shocks—2001 terrorist attacks, 2002 tropical storm, 2004 Hurricane Ivan, 2005 Hurricane Emily, the commodity price shock in 2008, and global financial crisis followed by a substantial economic slowdown thereafter. While “import booms” tend to be strongly positively correlated with the output gap, indirect taxes could be more closely related to import gap suggesting that using the output gap as the only proxy for the cycle could distort the cyclical component of the fiscal balance.3 In addition, grant flows have in general been high in recent years making it difficult to disentangle whether the average level of grants (“new normal”) has increased or whether it is purely in response to the series of shocks the country has experienced.4

This annex constructs alternative estimates of the structural fiscal balance. The next section briefly outlines the methodology for estimating the structural fiscal balance, Section III provides estimates from this approach and Section IV concludes with a sensitivity analysis.

B. Approaches to Estimate the Structural Primary Fiscal Balance

Staff estimates the structural primary fiscal balance using two approaches: (i) the potential output (IMF approach) and (ii) a hybrid of potential output and import norm (Box A1). The analysis focuses on the structural fiscal balance including grants given that spending tended to be adjusted in response to grants but it also illustrates the path excluding grants.5

Estimating the Structural Fiscal Balance IMF approach:

IMF approach:

FBsYp=(YpY)ε(RYp)+(GYp)(EYp)

Hybrid approach:

FBsYp=(YpY)ε2(ROYp)+(ZpZ)ε1(RIYp)+(GYp)(EYp)

where FBs is the structural fiscal balance; R is structural revenues (excluding grants); G is grants and E is expenditures net of one-off factors; Y and Yp are actual and potential output, RI are indirect taxes, RO are other respect to the output gap, and elasticities of indirect and other taxes to the output and import gaps, respectively.1 revenues, Z are imports, Zp is “normal” level of imports and ɛ, ɛ1 and ɛ2 capture the elasticity of revenue with respect to the output gap, and elasticities of indirect and other taxes to the output and import gaps, respectively.

  • The IMF approach using potential output derives an estimate of the structural fiscal balance by removing the impact of the business cycle on fiscal variables. The business cycle is derived by estimating the output gap using the Hodrick-Prescott (HP) filter, the most commonly used approach,6 and the structural balance is computed by applying elasticities of the fiscal variables to the output gap. Following the IMF’s Fiscal Affairs Department (FAD) methodology,7 revenues (excluding grants) are assumed to have unit elasticity, as they tend to be positive correlated to the business cycle, while expenditure and grants are assumed to have zero elasticity (i.e., not responsive to the business cycle).8 In addition, the elasticities of revenue and grants for Grenada during 1991-2012 are estimated to see how the results may differ. The elasticities are estimated using ordinary least squares (OLS) and quantile regression—since the elasticity could be a function of the size of the fiscal variables. The findings for the revenue elasticity using the OLS approach indicate that it is 0.13 and highly significant but insignificant for expenditure during the same period. However, quantile regressions point to an elasticity of about 0.6 and is highly significant for all quantiles for the revenue variable but only significant for the 75th quantile for the expenditure variable, with highly significant coefficients. Sensitivity analysis is conducted applying these elasticities and also using alternative approaches to estimate the output gap, since the HP filter will only provide information about the current output gap if the gap is “driven” by output growth in the previous period (i.e., granger-caused and not exogenous) and has weaknesses at the end of the sample.

The hybrid approach derives an estimate of the structural balance by removing the impact of the import cycle on indirect taxes and the output gap on other fiscal variables.9 This is especially relevant for Grenada where about two thirds of total revenues (excluding grants) come from indirect taxes (taxes on goods and services and international trade) (Figure A2). Unit elasticities are then applied for both components of revenue as country specific estimates are hindered by the lack of sufficient degrees of freedom and structural break in the relationship overtime.10 A key challenge is appropriately estimating an import boom. We assume that the share of imports in 2018 (medium term stable level) provides an indicator of imports at “normal” levels.11

Figure A2.
Figure A2.

Composition of Revenues

(In percent of revenues, excluding grants)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Sources: Country authorities and staff estimates.1/ Revenues (excluding grants) = Capital + direct tax + indirect tax + non-tax revenues. Capital revenues are the residual. Not shown but very small.

Transitory shocks and measures are excluded from the structural balance. Though it is difficult to identify, track and remove one-off operations and accounting events, this adjustment is in line with the consensus in the literature.12 The European Commission, for example, has established guidelines to limit discretion in the calculation of one-off adjustments (EC 2006). According to these guidelines, to be considered as one-off, a measure must: (i) have a significant impact in the budget (at least 0.1 percent of GDP), (ii) have a limited temporary impact and (iii) be non-recurrent. In Grenada, the EC$3 million (0.2 percent of GDP) that was included as net lending in 2005 is considered one-off. Ideally, large one-off grants and grant-financed expenditure could be adjusted but this is difficult since they have been very volatile in Grenada, with many years of above-average financing (e.g., 2003-2006; 2008-2011) coupled with periods of drops in financing (e.g., 2007) making it difficult to decompose the trend from the cycle.

The stance and thrust of fiscal policy is assessed using the fiscal impulse.13 In essence, this provides a measure of the initial impact of discretionary fiscal policy on aggregate demand and is typically defined as the change in the structural balance, with a positive (negative) number implying expansionary (contractionary) effect. In Box A2, equation (3) illustrates the methodology; it augments the typical definition to address biases that occur using the potential output gap approach.

Sensitivity analysis is needed given the difficulties in measuring potential output and capturing the import norm. This includes different approaches to capture the gaps and also country-specific estimates of revenue elasticity to these gaps (see Section IV).

Deriving the Fiscal Impulse

Using the IMF WEO methodology, Heller et al. (1986) calculate a measure of the fiscal impulse

F1=(ΔE-eoΔYp)-(ΔR-roΔYp)-(ΔG-goΔYp)(1)

where: FI is the absolute measure of fiscal impulses, eo, ro and go are the shares of government spending, revenue (excluding grants) and grants to nominal GDP in the base year, respectively, E, R, G, Y and Yp are all in nominal prices, and Δ is the difference operator.

However, the fiscal impulse measure is sensitive to the choice of the base year and measurement of potential GDP. Modifications are made to avoid the base year bias. Here the impulse is determined by reference to the preceding year’s budget balance:

F1=(ΔE-nEt-1)-[(ΔR-(ΔY/Yt-1)Rt-1)-(ΔG-nGt-1)(2)

where

n=ΔYpYp,t1

article image
is the growth rate of potential output

C. Empirical Estimates of the Structural Balance

Using the “IMF approach”, the structural primary fiscal deficit declined between 2006 and 2012. (Figure A3).14 There was a widening in this deficit during the initial years of the two programs (2007 and 2011, respectively) which coincided with the programs going off-track, as fiscal policy became expansionary (indicated by the positive fiscal impulse). While the structural primary deficit since 2006 had been much higher than the pre-program period, the fiscal adjustment and grants could have been lower in the absence of a Fund supported program. Examining the structural primary fiscal balance abstracting for grants (i.e., excluding grants from revenue), suggests that there was a large deterioration in the structural primary balance, excluding grants in 2006 and 2008 mainly driven by a surge in capital spending and recurrent spending, respectively.15 Since there was a substantial increase in capital spending related to the hurricanes, it is probably not realistic to remove grants without making some adjustment for expenditure. As noted earlier however, grant flows have in general been high in recent years making it difficult to determine the “normal” level. Therefore, using the primary balance including grants may be a better measure.

Figure A3.
Figure A3.

Grenada: Fiscal balances

(In percent of GDP, unless indicated otherwise)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: Country authorities and staff estimates.1/ A positive (negative) fiscal impulse implies expansion (contraction). Fiscal impulse is the augmented fiscal impulse

The structural primary fiscal deficit using the hybrid approach is much bigger (compared to the approach based on the output gap) particularly during 2005-2008 and reaches its peak in 2007 (Figure A4). It is broadly the same as the standard structural primary fiscal balance in 2009 as the output gap was practically closed and the imports were broadly in line with the “norm”.

Figure A4.
Figure A4.

Grenada: Structural Fiscal Balance and Import gap

(In percentof potential GDP, unless otherwise indicated)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: Country authorities and staff estimates

The fiscal impulse analysis suggests that fiscal policy was contractionary during most of the time the program was active (i.e., not off-track).16 Based on the IMF approach using potential output and computations specified in Box A1, fiscal policy was expansionary in 2006 and 2007 during the period the program went off-track but became contractionary during 2008-2010 as revenues plummeted and capital spending was cut sharply in 2009. The fiscal impulse remained contractionary in 2010 (Figure A5). Overall the bulk of the adjustment fell on expenditure, in particular capital spending. The fiscal impulse (particularly expenditure) contributed to a contraction in aggregate demand, adjusted for the business cycle. The biggest impact occurred in 2009 where it contributed to about a 50 percent decline in aggregate demand growth. For example, in 2009 aggregate demand growth was -6.6 percent and the expenditure impulse contributed to about half of this decline. However, the multiplier impact of fiscal policy on output growth in Grenada is low probably due to the large import component and high unsustainable debt (see Box 6 in the main paper).

Figure A5.
Figure A5.

Grenada: Fiscal Impulses

(In percent of Previous Year’s GDP) 1/

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

1/ (-) = fiscal withdrawal. The difference between the fiscal impulse and the expenditure and revenue impulses is the grant impulse.

D. Sensitivity Analysis

Sensitivity analysis shows that the direction but not the size of the change in the structural fiscal balance is robust to alternative specifications. Given the uncertainty surrounding the measurement of the potential output and normal level of imports as well as the elasticities, alternative estimates are examined.17 Using Grenada-specific estimates of elasticities for revenue and expenditure and the HP filter (to derive potential output gap) decreases the estimated structural deficit in 2006-2009 but increases it thereafter (Figure A6). The estimates are particularly sensitive to the approach taken to compute the output gap and suggest a much bigger reduction in the structural balance between 2006 and 2010 when we use the Christiano-Fitzgerald filter to derive potential output gap and apply unit elasticities. Specifically, applying the IMF suggested elasticities, the structural deficit declines by about 5.1 percent using the CF filter approach compared to about 3.9 percent using the HP filter. This reflects the somewhat larger output gap using the HP filter.

Figure A6.
Figure A6.

Sensitivity Analysis of the structural primary fiscal balance

(In percent of potential GDP)

Citation: IMF Staff Country Reports 2014, 019; 10.5089/9781484389362.002.A001

Source: Country authorities and staff estimates.

Appendix I. Grenada: Observance of Structural Conditionality in Fund-Supported Programs

article image

Language describing this measure differs somewhat in program documents for the three reviews at which it was assessed.

26 measures in 1st program.

7 met

9 met with delay

10 not met/dropped.

References

  • Bezdek, Vladimír, Dybczak, Kamil, and Ales Krejdl, 2003Cyclically Adjusted Fiscal Balance—OECD and ESCB Methods, Czech Journal of Economics and Finance, 53.

    • Search Google Scholar
    • Export Citation
  • Bornhorst, Fabian, Dobrescu, Gabriela, Fedelino, Annalisa, Gottschalk, Jan, and Taisuke Nakata, 2011,—When and How to Adjust Beyond the Business Cycle? A Guide to Structural Fiscal Balances, IMF Technical Notes and Manuals, April (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Bouthevillain, Carine, Cour-Thimann, Philippine, van den Dool, Gerrit, Hernandez de Cos, Pablo, Langenus, Geert, Mohr, Matthias, Momigliano, Sandro, and Mika Tujula, 2001, Cyclically-Adjusted Budget Balances: An Alternative Approach, ECB Working Paper No. 77 (Frankfurt: European Central Bank).

    • Search Google Scholar
    • Export Citation
  • Buti, M. and N. Carnot (2013), “The debate on fiscal policy in Europe: beyond the austerity myth”, Ecfin Economic Briefs 20, March.

  • Escolano, J., 2010, “A Practical Guide to Public Debt Dynamics, Fiscal Sustainability, and Cyclical Adjustment of Budgetary Aggregates,IMF Technical Note and Manuals 10/02

    • Search Google Scholar
    • Export Citation
  • European Commission, 2006, The State of Public Finances.

  • Fedelino, Annalisa, Ivanova, Anna, and Mark Horton, 2009,—Computing Cyclically Adjusted Balances and Automatic Stabilizers, IMF Technical Notes and Manuals, November (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hagemann, Robert, July 1999,—The Structural Budget Balance: The IMF’s Methodologyll, IMF Working Paper No. 99/95 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Heller, P.S., Haas, R.D., and A.S. Mansur, 1986, “A Review of the Fiscal Impulse Measure”, IMF Occasional Paper No. 44.

  • IMF, 1985, World Economic Outlook (Washington: International Monetary Fund), October

  • IMF, (2010) “Annex IIIin Montenegro: 2010 Article IV ConsultationStaff Report; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for Montenegro”, International Monetary Fund, May (Washington: DC).

    • Search Google Scholar
    • Export Citation
  • OECD, 1985, OECD Economic Outlook, OECD December.

1

The other ECCU members are: Antigua and Barbuda, Dominica, St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines, as well as Anguilla and Montserrat, which are overseas territories of the United Kingdom.

2

A country is considered to have longer-term program engagement when it has spent 7 or more of the last 10 years under Fund-supported financial arrangements without counting time spent under precautionary arrangements that remained undrawn (see, The Acting Chair’s Summing Up - Review of Ex Post Assessments and Issues Related to the Policy on Longer-Term Program Engagement http://www.imf.org/external/np/sec/pn/2006/pn0696.htm and Ex Post Assessment of Members with a Longer-Term Program Engagement, Revised Guidance Notehttp://www.imf.org/external/np/pp/eng/2010/022510a.pdf.

3

The World Bank defines upper-middle income economies as those with per capita GNI of between $4,086 and $12,615.

4

Grenada became PRGT-eligible in 1986. The other PRGT-eligible countries in the ECCU are Dominica, St. Lucia, and St. Vincent and the Grenadines. Grenada’s per capita GNI currently exceeds the IDA operational cut off by a factor of six, which is relevant for its graduation from PRGT-eligibility, but remains PRGT-eligible given the presence of serious short-term vulnerabilities, especially given very high public debt (see Box 3, and more generally, Eligibility to Use the Fund’s Facilities for Concessional Financing http://www.imf.org/external/np/pp/eng/2013/031813a.pdf.

5

The expansionary fiscal policy was mainly due to increased capital expenditure that was financed in large part by the issuance of international bonds, at relatively high interest rates (9.5 percent). The increased borrowing elevated interest payments (doubling in 2002 relative to 1999) remaining high until the debt restructuring in 2006. Box 6 discusses the fiscal multipliers.

6

The hurricanes resulted in damage estimated at 200 percent of GDP and severely affected Grenada’s main export earners, tourism and nutmeg crops which need to grow at least for 7-10 years until they can be harvested and 13 years to reach peak production levels.

7

Initially, purchases were made under the Fund’s Emergency Natural Disaster Assistance (ENDA) in 2004.

8

Since the outturn for 2013 is not yet known and the last ECF program expired in April 2013, the “assessment period” for the two programs covers 2006-2012, with the first program period covering 2006-2009 and the successor program covering 2010-2012.

9

As the furthest island from major air and cruise ship hubs (Miami and Fort Lauderdale) the high fuel cost caused some air and cruise ship companies to decrease or stop servicing Grenada.

10

The 2005 debt exchange resulted in a NPV reduction of 35 percent on debt service but no principal reduction and interest rates were set to rise from 1 percent in the first three years to 4.5 percent by 2012 (Box 3).

11

In 2011, the government issued EC$66.8 million in T-bills (3.1 percent of GDP) versus an average of EC$18 million issued in the previous three years. About half of these were issued in the regional government securities market, while the rest were sold in private placements to domestic entities and other investors.

12

Arrears are estimated at close to 5.6 percent of GDP at end-April 2013, of which 2.9 percent is to domestic suppliers and 2.7 percent external. In addition to external private creditors, the authorities are also in arrears to bilateral creditors.

13

Grenada’s general elections on February 19, 2013, resulted in a change of government. The incumbent National Democratic Congress (NDC) Party was defeated by the opposition New National Party (NNP) under the leadership of Dr. Keith Mitchell, who had governed the country during 1995–2008. The NNP won all 15 seats in parliament.

14

Two insurance subsidiaries of CL Financial Group had been offering deposit-like products: British American Insurance Company (BAICO) and the Colonial Life Insurance Company (CLICO). BAICO accounted for just over half of the exposure to policy and deposit holders in the ECCU (EC$2 billion; 15 percent of ECCU GDP). At the regional level, BAICO was estimated to have a negative net worth of some EC$0.8 billion (6 percent of ECCU GDP).

15

Credit unions and insurance companies also had written off their exposure to the CL Financial Group by the end of 2012.

16

A country poverty assessment was completed in December 2009 with assistance from the CDB.

17

Under the new architecture for low-income country lending, PRGF arrangements were converted into ECF arrangements on January 7, 2010.

18

Under the revamped lending policies introduced in 2009, the use of structural performance criteria was discontinued in all Fund arrangements, and conditionality attached to structural reforms takes the form of structural benchmarks which are monitored in the context of program reviews. See http://www.imf.org/external/np/pdr/fac/2009/032409.htm.

19

An adjustment was included to accommodate grants and higher than programmed financing from external creditors, including concessional financing and existing donor-financed projects.

20

The performance criterion on domestic arrears older than 60 days was revised in the Second Review of the first Fund-supported program from monitoring on a continuous basis to an outstanding stock basis at the specified test dates.

21

Technically, only one-third of the targets on the primary balance were met, but the target in the third review was only missed by a small margin and was treated as de minimis misreporting.

22

As envisaged under the 2006 program, capital spending was to be brought down to the historical average of 8 percent of GDP however by 2012 it had declined to about 5 percent of GDP.

23

Staff research suggests that fiscal multipliers are small in the ECCU. However, the multiplier for public investment could be as large as 0.6 (see IMF Working Paper, “Fiscal Multipliers in the ECCU” by Gonzalez-Garcia A. Lemus, and M. Mrkaic).

24

Social spending includes pensions, NIS contributions, grants and subventions, and transfers to statutory boards. Data were obtained from WEO.

25

With the help of CARTAC, GDP in current prices was revised upward in 2010 which significantly reduced the debt ratio, but it remained well above comfortable thresholds. The revisions also made it more difficult to monitor debt targets.

26

The Fund encouraged the authorities to undertake a feasibility study and examine options to include substantial equity participation from private investors to ensure the risks are shared across all parties.

27

The program included a number of adjustors including on grants, concessional financing from external creditors (excluding IMF and Petrocaribe), divestment proceeds and bank restructuring costs.

28

Controlling expenditure was made more complex since many items that were classified as capital outlays were recurrent spending with significant capital-formation type effects and the converse for recurrent spending. An EU report assessing public finance management in Grenada noted that some expenditures are classified as capital (e.g., tourism promotion) but should be regarded as current under strict GFS definitions, while some capital expenditure (purchases of tools and equipment) are treated as recurrent. EU (2009) Assessment of Public Financial Management in Grenada using the PEFA PFM Performance Framework, Final report, Project No. 2009/217871.

29

CARTAC was to put in place a comprehensive commitment control system by February 2011 under the successor Fund-supported program.

30

The review of conditionality in 2012 highlighted areas where further strengthening in implementation of underlying policies might be required, including: (i) keeping conditionality focused; (ii) enhancing risk diagnostics underpinning program design; (iii) considering macro-social issues in IMF-supported programs; (iv) enhancing program ownership and transparency; (v) leveraging economic surveillance to increase contingency planning; and (vi) improving partnerships with other institutions. See: http://www.imf.org/external/np/sec/pn/2012/pn12109.htm.

31

GrenadaRequest for Three-Year Arrangement under the Poverty Reduction and Growth Facility, http://www.imf.org/external/pubs/ft/scr/2006/cr06277.pdf, page 3.

32

GrenadaRequest for Three-Year Arrangement under the Poverty Reduction and Growth Facility, http://www.imf.org/external/pubs/ft/scr/2006/cr06277.pdf, page 16.

33

Structural benchmarks aiming at amending or repealing the relevant legislation were set for the first, second, and third program reviews, and repeatedly missed before efforts in this area appear to have been abandoned.

34

The authorities appear to have eventually developed an action plan to improve their ranking in the Doing Business indicators, but this was never adopted as IMF program conditionality.

35

The CDB also on-lends funds committed by the Inter-American Development Bank (IADB).

36

Doing Business Indicators must be considered with caution, owing inter alia to methodological and coverage changes that complicate interpretation of changes of time, the subjective nature of some components, and the uncertain link between the indicators and long-term growth prospects. We have used the Indicators to provide a broad sense of trends in Grenada’s investment climate, relative to regional peers.

37

This included financing from the Caribbean Development Bank (CDB) under the Policy Based Loan, access to the V-flex facility from the EU, and project loans from the World Bank and CDB.

38

Grenada’s small population necessitates a relatively small public service, and key economic agencies are as a result thinly staffed. In addition, like other Caribbean countries, Grenada may suffer from a brain drain.

39

At the time of the first review this was discussed as one of the ways the ministry would improve control of spending, however, in the second review the fiscal targets were missed and a circular was issued as a prior action for the next review (see earlier discussion). The 2007 Article IV discussed the possibility of setting up a contingency fund to help buffer against future shocks.

40

As noted in the recent growth and jobs paper, “Jobs and Growth: Analytical and Operational Considerations for the Fund,” March 2013, the “growth diagnostics” approach of Hausmann, Rodrik, and Velasco recommends that countries “need to figure out the one or two most binding constraints on their economies” and that a focus on lifting this is preferable to having a “laundry list of needed reforms,” many of which may not be “crucial to [the] country’s growth potential” (Hausmann, Rodrik, and Velasco, 2006).

41

In December 2012, Grenada’s Caribbean Growth Forum started and has produced several reports which focus on 3 main areas: 1) logistics and connectivity (including energy costs, regional collaboration); 2) investment climate (private sector development and investment promotion); 3) skills and productivity in the government’s priority sectors (tourism, agriculture, education, health, and energy). The Caribbean Growth Forum is a joint initiative by the Compete Caribbean Program, the Inter-American Development Bank, the World Bank, and the Caribbean Development Bank to facilitate an action oriented dialogue around key policy reforms to support sustainable and inclusive growth in the Caribbean.

42

See the World Bank’s World Development Report 2013: Jobs and Jobs and Growth: Analytical and Operational Considerations for the Fund http://www.imf.org/external/np/pp/eng/2013/031413.pdf.

43

Staff participated in Growth Forum exercises and a joint Growth Conference in 2013.

1

The focus of this analysis is on examining the sensitivity of the estimates derived using the IMF methodology (used in the main report) to changes in the approach used to compute the output gap and to different estimates of elasticities.

2

The Fund arrangements with Grenada applied nominal targets on the primary balance excluding grants (with adjustors to increase flexibility to changes in grant and other financing) due to difficulties in estimating the structural fiscal balance and the need to reduce the high and unsustainable public debt.

3

This effect is not fully captured with the standard methodology that adjusts only for the output gap and not for the external imbalance.

4

Grenada received substantial grants to help it mitigate the impact of a series of adverse shocks; this coupled with the size and duration of these grants raises a question about whether the “normal” level of grants has increased.

5

The correlation between capital grants (grants tend to be capital grants) and capital expenditure is 0.5. In general, grants are excluded in the analysis of the fiscal stance since they do not capture policy effort. However, this assessment may not be fully appropriate in Grenada since spending was adjusted in the face of shocks.

6

The IMF does not have an “official” method to compute potential output. Each country desk chooses the method that best reflects the country particularities. However, the common approach utilized has varied over time and historically relied on a production function, with the underlying assumptions varying across countries. In recent years, the HP filter has been used. This analysis focused on the most recent approach as outlined in Fedelino, A. et al., (2009). See De Masi (1997) for a detailed review of the historical approaches to the estimation of potential output.

7

See Fedelino, A. et al. (2009), for more details.

8

The IMF methodology does not specify how to treat grants. The approach used here assumes grants are exogenously determined similar to that used in previous estimates for Grenada (see http://www.imf.org/external/pubs/ft/scr/2010/cr1014.pdf, Table 3).

9

The literature has noted that a number of real activity variables (e.g., consumer spending, industrial output and wage bill) may fluctuate differently from output and suggested shifting the structural balance focus out of the output gap to cyclical movements in other variables that are more closely linked to the fiscal aggregates (e.g., Bouthevillian et al, 2001; Bezdek et al, 2003, and IMF, 2010.

10

Unit elasticity for output gap and import boom draw on the IMF methodology due to lack of sufficient data on decomposed revenue data. Econometric analysis yields somewhat different estimates but this could reflect the limited degrees of freedom. Estimates for Grenada suggest that the elasticity of revenue excluding grants to the import gap is about 0.3 (-1.3 using quantile regressions) and significant but the elasticity of indirect and other taxes since 1999, as decomposed historical data before this is missing, is -0.2

11

The WEO projections for Grenada in 2018 are applied. These are typically viewed as the steady state values.

12

See Borhorst et al (2011), Fedelino et al. (2009) and Oreng, 2012 for a discussion.

14

As noted in the IMF technical guidance notes (Fedelino et al., 2009), scaling is an important issue for the structural balance. The structural balance estimates are presented as a share of potential ouput as this improves the comparability of results and is the way most researchers report their estimates. However, our results are broadly insensitive to scaling.

15

While the structural balance excluding grants provides an indicator of policy effort—since grants are provided by donors—to the extent that expenditure reflects grants, excluding grants could bias the findings.

16

For brevity, the analysis illustrates the fiscal impulse using the potential output gap approach.

17

The focus of this analysis is on the sensitivity to the estimates derived using the IMF methodology to changes in the approach used to compute the output gap and to different estimates of elasticities.