This Selected Issues paper assesses the fiscal risk impact of various events in Tonga and describes the challenges implied for fiscal management. The paper discusses that the Tongan government has made various efforts to limit the budgetary implications of the recent economic and political shocks, including pursuing revenue administration efforts, cutting other expenditures, and seeking donor assistance. The paper examines rising household indebtedness in Tonga. Within the framework of the life-cycle hypothesis, the paper finds that Tonga’s relatively young demographic distribution tends to be associated with higher level of aggregate household debt.

Abstract

This Selected Issues paper assesses the fiscal risk impact of various events in Tonga and describes the challenges implied for fiscal management. The paper discusses that the Tongan government has made various efforts to limit the budgetary implications of the recent economic and political shocks, including pursuing revenue administration efforts, cutting other expenditures, and seeking donor assistance. The paper examines rising household indebtedness in Tonga. Within the framework of the life-cycle hypothesis, the paper finds that Tonga’s relatively young demographic distribution tends to be associated with higher level of aggregate household debt.

I. Challenges to Sustainable Fiscal Management in Tonga1

A. Background

1. Tonga has faced a number of fiscal shocks over the past two years. Following a lengthy civil service strike in September 2005, which lasted for six weeks, the government agreed to a wage increase of 60–80 percent for all public employees. This wage award posed serious difficulties for fiscal management, and the government subsequently downsized the civil service by 18 percent in 2006 as a mitigating measure. To avoid social unrest, the retrenchment program was carried out on voluntary basis.

2. Political uncertainty has also increased the country's vulnerability to shocks. The civil service strike became directed against the government, and was joined by various segments of society expressing their dissatisfaction with the political regime and calling for greater democracy. In response, the government committee considering the political reform process proposed the introduction of parliamentary democracy, but the government's decision to defer the desired constitutional changes sparked riots in November 2006.2 This disturbance caused serious damage to the central business district in Nuku'alofa, with grave risks for economic stability.

3. This chapter assesses the fiscal risk impact of these events and describes the challenges implied for fiscal management. The Tongan government has made various efforts to limit the budgetary implications of the recent economic and political shocks, including pursuing revenue administration efforts, cutting other expenditures, and seeking donor assistance. With these efforts, the near-term risks have been contained thus far, with the fiscal deficit kept within manageable levels. However, medium-term challenges persist including reforming the budget process and other measures to place the fiscal accounts on a more solid footing.

B. September 2005: Public Sector Wage Settlement

4. The public sector wage settlement imposed a serious fiscal burden on the government. As a result of the wage settlement, Tonga's already large civil service wage bill was raised well above levels elsewhere in the region. The agreed wage increase of 60–80 percent reversed the government's efforts in 2003/04 and 2004/05 to bring the wage bill down to comparable levels with regional peers, causing additional wage expenditure of an estimated 2¾ percent of GDP in 2005/06. The full wage increase, as well as deferred payment from 2005/06, was implemented in 2006/07, hiking the wage bill to above 18 percent of GDP.3 For 2007/08, despite the elimination of all one-off payments related to the settlement, the wage bill is projected to remain over 15 percent of GDP, substantially higher than the regional average of 10½ percent. The number of public employees also remains high in Tonga, even after the retrenchment program, accounting for 3.6 percent of the total population, compared to an average 2.6 percent in middle-income countries.

Developments in the Size of Public Services

article image

Excludes redundancy payment.

Staff projections.

Data for 2002–04.

5. Historical experience suggests the lack of proper planning of the public service with an orderly wage-setting structure has resulted in the rising real wage bill over time as well as excessive variability. The number of full-time civil servants went up by 60 percent between 1981 to 1995, bringing the size of Tonga's public sector work force up to 5 percent of the total population. Reflecting this expansion, the wage bill continued to increase until the mid-1990s and reached about 15 percent of GDP by 1995/96. Moreover, civil service wages were reviewed on a five-yearly basis, and a major review of public compensation in 1989/90 that awarded a 50 percent wage increase over two years, accelerated the rising trend of the wage bill. Another review in 1999/2000 also resulted in similar outcome, a 20 percent increase for cost of living allowances. As a result, the wage bill continued to increase through early-2000, risking fiscal sustainability.

uA01fig01

Wage Bill Development 1/

(In percent of GDP)

Citation: IMF Staff Country Reports 2007, 298; 10.5089/9781451837490.002.A001

1/Excludes redunduncy payment in 2005/06. Shadows indicate large wage adjustment during the periods.

6. Similar to other small island countries, retaining qualified staff in the public sector has been the government's key challenge. Due to a relatively flat wage scale, recruiting and retaining qualified staff has been a serious concern in Tonga, raising the need for comprehensive reform of the public sector wage structure. Supported by the Asian Development Bank, the government launched the Economic and Public Sector Reform Program (EPSRP) in 2003 with the aim of “rightsizing” the public services. Under the program, wage and salary increases were suspended, and recruitment to the public service was tightly restricted to priority positions. The EPSRP also included plans for giving differential salary increases based on performance and levels, as well as adopting a more conventional wage-setting approach.4 Reflecting these plans, the salary reform package approved by the government in 2005 proposed a significant salary increase for high level officials, while allocating relatively small increases for other public employees (Box 1). However, this proposal had not been widely discussed with public service unions, resulting in widespread dissatisfaction among civil servants especially among lower levels, and caused the civil service strike in July 2005. As a result of the wage settlement following the strike, the wage levels in the public sector are currently higher than those in the private sector, especially for workers in lower levels.5

Proposed Salary Reform Package

Based on recommendations made by the Higher Salaries Review Committee and a review conducted by the AusAID, the PSC proposed a comprehensive reform of civil service salary structure in 2005. The package included the following key recommendations.

  • Grant an increase in aggregate salary costs of about 13 percent, but with the increase applying to an individual depending on a full assessment of the work value;

  • Assess work value of positions against objective criteria;

  • Introduce an examination process of claimed anomalies in the system by the head of each department and agency; and

  • Reduce the number of salary classifications and the number of increments within grades, and eliminate salary overlaps.

7. The September 2005 wage settlement yet again highlighted the urgent need for an orderly wage-setting structure. While large wage increases are not uncommon in the Pacific Island Countries, the uncertainty over the size of the award as well as the risk of civil service strikes stemming from a five-yearly wage review continues to complicate fiscal management (Box 2). As further increases in the wage bill are clearly not affordable to the government, adopting an orderly wage-setting structure, including through more frequent review and design of appropriate benchmarks to guide wage awards, should be considered. A performance-based salary structure also remains critical to enhance the efficiency of the public sector through motivating and retaining qualified staff. In introducing such a system, moreover, care should be taken in designing the government's communication policy to keep civil servants fully informed, thereby perhaps increasing their receptiveness to reforms.

Experiences of Wage Increases in the Pacific Island Countries

Orderly wage setting has been a common challenge for a number of Pacific Island Countries (PICs). Due to large one-off wage settlements, often following extended periods of wage freeze, a number of PICs have experienced substantial increases in wage and salaries as a percent of GDP, as well as large variability in the real wage bill. Recent public sector wage increases in the PICs include the following:

  • Fiji: After months of negotiations over pay rise as well as the wage setting structure in 2004 and 2005, the dispute between the Public Service Commission and civil service unions went to arbitration, and the arbitrator reached a decision at end-2005. Civil servants were granted a wage increase about 10 percent in 2006, and, in addition, received one-off payments for retroactive pay adjustment. As a result, the wage bill rose by 15½ percent (y/y) in 2006, reaching 12 percent of GDP. More recently, however, the interim government has cut wages back by 5 percent in a revised budget adopted in March 2007.

  • Kiribati: In the 2005 salary review, public employees were granted a wage increase of an average 8.3 percent after a five year pay freeze, while the number of public employees was also increased by 1.3 percent. As a result, the total wage bill rose by 9.6 percent (y/y) in 2006, bringing the wage bill ratio to 43.8 percent of GDP.

  • Marshall Island: Wage and salaries spending has continued to rise since FY2001, due to increases in the number of civil servants and the harmonization of wage levels in the Ministry of Education. The wage bill increased by 58 percent from FY2001 to FY2005, reaching about 30 percent of GDP.

  • Samoa: In the 2005 salary review, after consecutive years of a wage freeze, the government decided to increase public sector wages by 42 percent over three years, starting from July 2005. With the second pay rise, the wage bill is now budgeted about 8.5 percent of GDP for FY2006/07, compared to 7.2 percent in FY 2004/05.

  • Vanuatu: After a 10 year wage freeze, the government agreed to increase public service wages by 20–25 percent, starting from July 2006. With this wage increase, the wage bill is expected to reach over 14 percent of GDP in 2007, compared to 10½ percent in 2005.

C. June 2006: Public Sector Downsizing

8. To mitigate the budgetary implication of the wage settlement, the government embarked on a public sector downsizing in March 2006. With the urgent need to contain wage spending, the government announced plans to reduce the size of the public service by about 20 percent within a three month period. To avoid further strikes, this downsizing program was required to be conducted on a voluntary basis, while making careful efforts to avoid potential disruption of service delivery. The government also intended this program to be a targeted redundancy based on skills and age, avoiding adverse selection problems. Helped by the relatively generous redundancy package including bonus payments, which were raised in late-May to increase the attractiveness of the package, the government succeeded in cutting 18 percent of the civil service by end-June 2006, reducing the overall size of public service to 3,600. The redundancy package cost TOP 24.5 million (5.4 percent of GDP) and was mainly financed by the government running down its domestic cash balances, depleting its fiscal cushion against future unexpected expenses. Currently, the government is imposing a strict limit on new hiring, keeping the number of workers at the current levels.

Tonga: Public sector retrenchment program

article image

Initial TOP 1,000 was increased to TOP 1,500 to attract more applicants.

9. The scale of Tonga's downsizing was unusually large. A comparison with crosscountry experiences of public service downsizing suggests that Tonga has undertaken a large cut in the size of public service in a remarkably short time period, substantially reducing the wage bill, albeit at the expense of a generous redundancy package. In fact, Tonga achieved a reduction in the number of public workers equal to 0.8 percent of its total population in three months, while other countries which have undertaken relatively large downsizing carried out their retrenchment programs over many years, averaging a period of five years. Although the total cost of severance payments also stands out in Tonga, at five times the cross-country average, the wage bill was reduced by over 3 percent of GDP, which is substantially higher than in other countries. By these measures, the downsizing program in Tonga was a significant achievement. Nevertheless the relative success of retrenchment programs cannot be simply evaluated by cost savings alone.

Country Experiences of Civil Service Downsizing

article image
Source: Tongan authorities; and Haltiwanger and Singh (1999).

Wage bill savings from separations minus cost for rehires and new hires.

Wage bill savings exclude potential cost for rehires and new hires.

10. Such a substantial downsizing within a short time period raises concerns about potential private and social costs. Although many of the relevant private and social costs are difficult to quantify, cross-county experience suggests that programs with a targeted multidimensional approach tend to be partly paid off by productivity gains and lower adjustment costs, even if they are initially expensive. Governments are less likely to resort to rehiring if retrenchment programs are appropriately targeted, and combined with assistance for workers reallocation.6 Although the retrenchment program in Tonga was intended to be a targeted redundancy, there appear to be significant risks of rehiring over time, as several public workers with valued skills accepted the redundancy package, and this has started to disrupt public service delivery.7 Moreover, given the short-time period over which the program was carried out, the government could not have provided sufficient assistance, such as necessary training to retrenched workers. The preliminary results of a follow-up survey conducted by the government almost a year after the event indicated that there were many retrenched workers still seeking employment, and that the majority of them needed assistance, such as job advisory services, training, and contact lists.8

D. November 2006: Civil Unrest

11. Government revenue collection appears to be affected by the unrest. Although estimating the revenue impact of the riots is difficult, monthly data for revenue collection clearly indicate an immediate revenue shortfall in the following months. Collection of consumption taxes and the customs tariff, the major revenue sources in Tonga, declined by 35 percent in November and December 2006, compared with the same period in 2005. Although part of these losses seem to have been offset by improved revenue collection in subsequent months, the fiscal room for exceptional spending on reconstruction and business support is limited: with the serious fiscal challenges the government faced in past years, the fiscal cushion has already been almost fully depleted after the redundancy payment, while the wage increase is still imposing an additional fiscal burden on the government.

uA01fig02

Consumption Tax Collection

(In millions of Pa'anga)

Citation: IMF Staff Country Reports 2007, 298; 10.5089/9781451837490.002.A001

uA01fig03

Customs Tariff Collection

(In millions of Pa'anga)

Citation: IMF Staff Country Reports 2007, 298; 10.5089/9781451837490.002.A001

12. Substantial business losses caused by the unrest have led to demands for restitution from the government. During the riots, buildings in the central business district in Nuku'alofa were burned, and small shops looted. Such direct losses from the unrest have been estimated at TOP 123.5 million (26.7 percent of GDP) by the government. Job losses were also substantial, accounting for 2.0 percent of economy-wide employment. Calls for government support for reconstruction and business recovery have become strong, especially given the political nature of the unrest. Nevertheless, the majority of affected businesses showed remarkable resilience to the disturbance and resumed their business operations after just a few weeks of disruption, relocating to remaining buildings, or to their warehouses.

Estimated Business-related Losses 1/

(In millions of Pa'anga; unless otherwise indicated)

article image
Source: Tongan authorities.

Estimation is based on reported damages as of December 2006.

13. The government announced a business assistance scheme in March 2007. The scheme announced is mainly for assisting private sector recovery through short-term interest rate subsidies as well as business advisory services. It also covers businesses which were indirectly affected, providing a small amount of cash grants as compensation for the loss of business opportunities. Financing for those schemes is expected to be provided by donors, and no additional fiscal expense is expected. However, some members of the business community regard the assistance scheme as insufficient for full business recovery, due to the strict application requirements, as well as the limited magnitude of the financial assistance on offer.

November 2006 Riots: Business Recovery Scheme

article image

14. To accelerate progress with reconstruction, the government is seeking access to large-scale financing from a donor. The expected reconstruction activities have yet to fully materialize, and the government has recently started to consider the possibility of financing reconstruction through external borrowings of about 20 percent of GDP. The rebuilt properties would be leased to private businesses. However, the loan under the negotiation is likely to entail most material input and labor being procured from the donor country, which would bring limited benefit to the domestic economy, while such a large loan itself involves substantial risks of future debt burden to the government. Demand for those new properties is also uncertain, as many of these affected businesses have already relocated to areas outside of the central business district.

E. Challenges Ahead

15. Keeping the fiscal accounts on a sustainable track is the key priority. Tonga has managed recent fiscal shocks with some success, but significant challenges still remain over the medium-term. Although political uncertainty continues to make fiscal management difficult, restoring fiscal sustainability is key to avoid further economic deterioration. Key policy recommendations to achieve this objective include:

  • Moving to more orderly wage-setting system. More orderly wage setting, including performance-based salary structure, should be adopted to avoid further wage shocks, while providing incentives to qualified staff.

  • Ensuring the quality of public service delivery. The quality of public service delivery after the retrenchment program should be carefully monitored. Resisting pressures of rehiring or new hiring is also critical, which may also require provision of nonfinancial assistance to retrenched workers to find job.

  • Improved public communication. Given the political sensitivity of the public sector reform, wider communication and consensus-gathering effort with public groups could possibly help avoid future social disturbance.

  • Providing the environment for the private to play a more active role in reconstruction. Given concerns about fiscal sustainability, more efforts should be made to involve the private sector in reconstruction of the central business district.

References

  • Asian Development Bank, 2005, “Economic and Public Sector Reform Program (Loan 1904-TON[SF]) in Tonga”, Program Completion Report

  • Asian Development Bank, 2006, Tonga “Social and Economic Update and Pro-poor Policy Formulation”

  • Haltiwanger, John and Singh, Manisha, 1999, “Cross-country Evidence of Public Sector Retrenchment” The World Bank Economic Review, Vol.13, No.1 2366

    • Search Google Scholar
    • Export Citation
  • Kingdom of Tonga, Higher Salaries Review Committee, 2004, “Report to the Prime Minister on the Remuneration of Holders of Judicial Office, Senior Civil Servants, Senior Officers of the Armed Forces and the Police Force and the Pay, Pensions and Allowances of Ministers and Noble' Allowances”

    • Search Google Scholar
    • Export Citation
1

Prepared by Ayako Fujita (Ext. 3-7368).

2

Responding to the public call for greater democracy, the government established the National Committee of Political Reform in 2005.

3

In 2005/06, civil servants received 60 percent of the total wage increase for the year, and the payment of the rest was deferred to 2006/07. As a result, in 2006/07, they received the full 60–80 percent salary increase as well as the deferred payment from the previous year.

4

Under the EPSRP, in 2003, the government set up the Public Service Commission as a public service appointment and promotion body. The Higher Salaries Review Committee was also set up in 2003 to review public sector wage levels (Asian Development Bank, 2005).

7

Under the program, a number of teachers took redundancy, which caused a serious problem in several schools. As a result, strong demand for rehiring or new hiring teachers has started to emerge in recent months.

8

The final outcome of the redundancy support services survey is scheduled to be presented to the Cabinet in 2007.

9

Prepared by Sutapa Amornvivat (Ext. 3-8774).

10

Refer to Table 7 in the Staff Report for the 2007 Article IV Consultation.

11

In this paper, credit to the household sector includes personal loans and mortgage loans from all financial institutions; plus farm and none-farm loans for small household enterprises which borrow from Tonga Development Bank.

12

Unlike in most emerging economies in Asia, there does not appear to be a housing bubble in Tonga due in parts to the lack of secondary markets for housing and rigid land tenure system.

13

Following the convention of corporate balance sheet analysis, two types of measures to gauge the strength of household balance sheets are solvency and liquidity ratios. Due to the lack of a national database on disposable income or assets, solvency is measured by the ratio of outstanding household debt to GDP.

14

Interestingly, mature market economies have the highest debt to GDP ratio on average, suggesting a higher degree of financial liberalization and higher income levels. This is not the case for Tonga.

15

Non-bank credit sources include credit unions, community cooperatives, micro-financial institutions, credit card companies, non-institutionalized moneylenders, relatives, and friends.

16

There is no clear evidence that a rise in banking efficiency contributed to higher household credit in Tonga. The entry of foreign banks started in the 1970s with Westpac and two other banks in the 1990s; there has been no change in the past five years that could explain the credit boom.

17

Tonga Development Bank has recently shifted towards more commercially-oriented operations.

18

As of end-2005.

Tonga: Selected Issues and Statistical Appendix
Author: International Monetary Fund