Journal Issue
Share
Article

Mexico: Selected Issues

Author(s):
International Monetary Fund
Published Date:
August 2004
Share
  • ShareShare
Show Summary Details

II. Towards Sustained Debt Reduction: Mexico’s Fiscal Framework1

This chapter examines the recent experience with Mexico’s fiscal management tools in light of the need for further consolidation. It concludes that a strengthening of the framework could play an important role in fostering deficit and debt reduction. We first review evidence on the cyclical behavior of fiscal policy in Mexico, finding that it has generally been procyclical since the early 1990s. We then assess the performance of existing fiscal management tools from both a cyclical and longer-term perspective. Finally, we analyze the contribution that various fiscal rules could make towards achieving sustainable deficit and debt reduction. In particular, a medium-term fiscal framework, emphasizing broader measures of public liabilities, and including explicit debt-targets and expenditure growth ceilings, could play a useful role in the Mexican context. In addition to institutional reforms to strengthen expenditure control, expanding the tax base would reduce revenue volatility, strengthen automatic stabilizers, and enable the government to run sustained primary surpluses without undue expenditure compression.

A. Introduction

21. Despite significant improvements in Mexico’s fiscal position since the tequilacrisis, important challenges remain. While both the traditional and broad deficit measures fell in the second half of the 1990s, consolidation has stalled recently. The primary surplus remains low and non-oil balances have deteriorated, as high oil prices have offset weak nonoil tax performance. Moreover, the ratio of gross public debt to GDP edged up to almost 50 percent at end-2002, in spite of high oil prices. Public finances are still vulnerable to internal and external shocks, reflecting a narrow tax base, dependence on oil revenues, and relatively large public sector liabilities. The debt ratio could rise further over the medium term, assuming oil prices decline from current high levels, unless decisive consolidation measures are taken. Further deficit and debt reduction are key for reducing the vulnerability of public finances and creating scope for counter-cyclical policies.

22. This paper examines the experience with Mexico’s fiscal management tools in light of the need for further consolidation. It concludes that a strengthening of the framework could play an important role in fostering deficit and debt reduction. We first review evidence on the cyclical behavior of fiscal policy in Mexico, concluding that it has generally been pro-cyclical since the early 1990s. The following section assesses the performance of existing fiscal management tools—the “budget adjustors” and the Oil Stabilization Fund (OSF)—in responding to shocks, as well as the authorities’ medium-term framework, the Programa Nacional de Financiamento del Desarollo (PRONAFIDE). Finally, we analyze the contribution that various fiscal rules could make towards achieving sustainable deficit and debt reduction. In particular, a medium-term fiscal framework, emphasizing broader measures of public liabilities, and including explicit debt-targets and expenditure growth ceilings, could play an important role in the Mexican context. In addition to institutional reforms to strengthen expenditure control, expanding the tax base would reduce revenue volatility, strengthen automatic stabilizers, and enable the government to run sustained primary surpluses without undue expenditure compression.

B. Cyclical Behavior of Fiscal Policy in Mexico

23. Several studies have concluded that fiscal policy throughout Latin America has tended to be pro-cyclical, preventing effective debt reduction during economic upswings (WEO, 2002;WEO, 2003;Gavin and Perotti, 1997; Talvi and Végh, 2000). Contributing factors have been: the weakness of automatic stabilizers due to narrow tax bases and limited social security systems; credit constraints on government financing during economic downturns; large public sector debt burdens and rising debt servicing payments in downturns; weaknesses in fiscal management techniques; and political and institutional factors that weigh against public savings—and debt reduction—during upturns. Social expenditures and infrastructure investment tend to be the first casualties during economic crises, exacerbating income inequality and potentially constraining long-term growth prospects.2Calderón and others (2002) estimate that more than half of the total fiscal adjustment effort in Argentina, Bolivia, Brazil, Chile and Peru during the 1990s was achieved through infrastructure compression. As a result, long-run growth may have been lowered by 1 percentage point per year.

24. A recent World Bank study concluded that fiscal policy has also been procyclical in Mexico (World Bank, 2001). Moreover, the authors found that a discretionary fiscal contraction is associated with a decline in GDP within the following two quarters. As a result, pro-cyclical changes in the fiscal stance have amplified the magnitude of the cycle. Social spending also shows pro-cyclical tendencies. During the1994–95 crisis, targeted spending per poor person decreased by 24 percent—driven both by reductions in social spending and an increase in the incidence of poverty (Hicks and Wodon, 2001).

25. Using the primary fiscal impulse as a measure of discretionary fiscal policy illustrates the procyclicality of policy since the early 1990s (Figure 1).3 Between 1991 and 2002, there was a significant, positive relationship between the primary impulse and changes in the output gap, with a correlation coefficient of 0.9. The factors underlying this tendency can be determined by decomposing the primary fiscal impulse into its components and estimating the relationship between changes in these fiscal variables and the cyclical component of GDP in a series of univariate regressions, using annual data for 1991 through 2002 (Table 1).

Figure 1:Output Gap and Fiscal Impulse in the 1990s

Table 1.Fiscal Variables and the Cycle, 1991-2002
Relationship to changes in the output gap

(as a percentage of potential GDP)
Betas 1/Correlation coefficients
Primary fiscal impulse0.35***0.90
Adjusted budgetary revenue 2/3/0.070.41
VAT revenue 2/0.00-0.20
Income tax revenue 2/0.07*0.52
Oil revenue (excl. export revenue) 2/-0.03-0.20
Other revenue 2/0.030.27
Total primary spending 4/5/0.42***0.90
Current spending 4/0.32***0.78
Capital spending 4/0.10**0.67
Memorandum items
Social spending 4/0.17***0.79
Interest spending 4/6/-0.46**-0.73
Source: SHCP and staff estimates.

Slope coefficients from a series of regressions of the change in fiscal variables on the change in the output gap.

Changes, as a percent of GDP.

Excludes nonrecurring revenue and oil export revenue.

Changes, as a percent of potential GDP.

Includes PIDIREGAS spending and all PSBR adjustors except financial requirements of development banks.

Includes interest costs of bank restructuring.

Note: ***, **, and * denotes significance on the 99 percent, 95 percent, and 90 percent level, respectively.
Source: SHCP and staff estimates.

Slope coefficients from a series of regressions of the change in fiscal variables on the change in the output gap.

Changes, as a percent of GDP.

Excludes nonrecurring revenue and oil export revenue.

Changes, as a percent of potential GDP.

Includes PIDIREGAS spending and all PSBR adjustors except financial requirements of development banks.

Includes interest costs of bank restructuring.

Note: ***, **, and * denotes significance on the 99 percent, 95 percent, and 90 percent level, respectively.

26. These results confirm a significant and positive relationship between the primary impulse and changes in the output gap. On the revenue side, the ratio of income tax revenues to GDP has a positive relationship with the cycle, but the low coefficient confirms their limited stabilizing role in Mexico. The ratio of VAT revenues to GDP is found to be insensitive to the cycle. Oil revenue is found to move slightly counter-cyclically. Primary expenditures, in contrast, are highly correlated with the cycle and appear to drive the overall procyclicality of fiscal policy. Current expenditure and social spending appear to be more sensitive to changes in the output gap than capital expenditure—in contrast to the experience elsewhere in the region that capital outlays tend to be disproportionately compressed during crises. Interest expenditures, on the other hand, are negatively related to changes in the output gap, presumably reflecting a negative relationship between sovereign yield spreads and Mexican activity.

27. Thus, fiscal policy in Mexico appears to be procyclical—or “leaning with the wind”—for a number of reasons. First, automatic stabilizers are weak, given the small tax base, and absence of an extensive social security system. At the same time, primary expenditures are highly procyclical. This procyclicality seems to be associated with countercyclical movements in debt-servicing costs. As Table 1 shows, the coefficients on primary expenditures and interest payments are roughly offsetting, indicating that overall expenditures have been little affected by the cycle. But as interest payments have fallen in upswings, the savings have been used to finance higher primary spending, and the reverse has occurred in downturns. This is consistent with the general difficulty observed in Latin America of containing political pressures to spend windfall gains (WEO, 2003).

C. The Budget Adjustors and Oil Stabilization Fund: Evolution and Impact

28. Two important fiscal management tools exist in Mexico to control budget execution and ensure saving of windfall revenues—at least in principle: automatic budget adjustors to deal with unforeseen fluctuations in revenues; and an Oil Stabilization Fund (OSF) designed to smooth the effects on the budget of short-term fluctuations in oil prices. As this section demonstrates, however, neither has been particularly successful in containing spending pressures during economic upturns or periods of high oil prices.

29. Since 1998, annual budget laws have included so-called fiscal adjustors. These adjustors imply pro-cyclical expenditure cuts when revenues fall short of budget estimates, and counter-cyclical deficit reduction when revenues exceed expectations. The adjustors have evolved since their introduction (Table 2). While written into annual budget laws, the government retains discretion in applying the adjustors. Ceilings on new net indebtedness are legally binding, though, and provide a limit to the overall deficit that can only be modified by congress.4

Table 2.Mexico: Annual Budget Adjustors
Response to Higher Revenues 1/Response to Lower Revenues
19981 percent of excess tax revenues over the first 9 months of the year to be spent on social and rural development, and road maintenance. The remaining 99 percent of excess revenues and all excess revenues of the last 3 months of the year to be saved.Shortfalls in overall revenues in excess of 1 percent of tax revenues to be matched by spending cuts. Congressional approval required for cuts in excess of 10 percent of original tax revenues.
19991 percent of excess tax revenues over the first 9 months of the year to be spent on social and rural development, and road maintenance.

The remaining 99 percent of excess revenues and all excess revenues of the last 3 months of the year to be saved.
Shortfalls in overall revenues in excess of 1 percent of tax revenues to be matched by spending cuts. Congressional approval required for cuts in excess of 10 percent of original tax revenues.
2000Of all excess revenue above 0.13 percent of GDP, 40 percent to be allocated to the Oil Stabilization Fund (OSF), and 60 percent to be used for debt amortization.Shortfalls in overall revenues in excess of 1 percent of tax revenues to be matched by spending cuts. Congressional approval required for cuts in excess of 10 percent of original tax revenues.
200133 percent of excess tax and oil revenue to be saved, 33 percent to be transferred to the Oil Stabilization Fund, and 34 percent to be spent on infrastructure investment in the South-Southeast, hydraulic infrastructure, and development projects in disadvantaged and oil-producing regions.Shortfall in oil nontax revenue (derechos) caused by a lower oil price of up to $1.5/bbl to be met through full expenditure adjustment. If oil price lower by more than $1.5, up to 50 percent of total resources in OSF can be used. Shortfalls in excess of this amount to be met through expenditure adjustment.

Any other shortfalls in budgetary revenues to be matched by spending cuts. Congressional approval required for cuts in excess of 5 percent of original tax revenues.
200233 percent of excess tax, nontax and oil revenue to be saved, 33 percent to be transferred to the Oil Stabilization Fund, and 34 percent to be spent on infrastructure investment in the South-Southeast, hydraulic infrastructure, and development projects in disadvantaged and oil-producing regions.All shortfalls in oil revenue (not only those caused by a lower oil price) can be met through up to 50 percent of total resources in OSF. Shortfalls in excess of this amount to be met through expenditure adjustment. Transitory article permitted almost full depletion of OSF in March 2002 (M x$8 billion).

Any other shortfalls in budgetary revenues to be matched by spending cuts in specified areas. Congressional approval required for cuts in excess of 5 percent of original tax revenues.
200325 percent of excess tax, nontax and oil revenues to be saved, 25 percent to be transferred to the Oil Stabilization Fund, and 50 percent to be spent on infrastructure investment in the States.Shortfalls in oil revenue can be met through up to 50 percent of total resources in the OSF. Shortfalls in excess of this amount to be met through expenditure adjustment. 2/ Any other shortfalls to be matched by spending cuts in specified areas.

Congressional approval required for cuts in excess of 5 percent of original tax revenues.
Source: Leyes de Ingresos y Egresos 1998-2003.

Net of higher “nonprogrammable spending”, such as federal revenue sharing and interest expenditure.

As resources in the OSF were almost completely depleted in 2002, this effectively implies full expenditure adjustment.

Source: Leyes de Ingresos y Egresos 1998-2003.

Net of higher “nonprogrammable spending”, such as federal revenue sharing and interest expenditure.

As resources in the OSF were almost completely depleted in 2002, this effectively implies full expenditure adjustment.

30. The adjustors were mainly designed to ensure that the “traditional” deficit does not exceed the budget target. Indeed, the Mexican authorities have been very successful at meeting these targets (Figure 2). The traditional deficit is much smaller than the broader “augmented” deficit concept used in most staff analysis, however, as it excludes large public sector liabilities associated with bank restructuring in the aftermath of the Tequila crisis, PIDIREGAS investment projects, and financial requirements of the development banks (Figure 3). For example, the 2003 budget envisages a traditional deficit of 0.5 percent of GDP, whereas the PSBR amounts to 3.4 percent of GDP. Since 2001, the authorities have published net public sector borrowing requirements (PSBR) along with the traditional deficit.

Figure 2.Mexico: Traditional Deficit Budget Targets and Outturns 1998-2002

(In percent of GDP)

Source: SHCP and staff estimates.

1/ Excluding Banrural

Figure 3.Mexico: The Traditional and Augmented Deficit Definitions

(In percent of GDP)

31. The character of the adjustors changed with the introduction of the OSF in 2000. Previously, oil revenues had been included in the overall revenue concept applied in the case of revenue shortfalls; they were excluded, however, from the definition of excess revenue, which was to be saved. In other words, oil revenues could compensate for non-oil tax revenue shortfalls, and windfall gains in oil revenues could be spent as long as there was not a shortfall in other revenue. In contrast, the original rules of the OSF were designed to provide a more robust cushion against oil price fluctuations: resources from the OSF could be used to offset low oil revenues when oil prices fell below a certain threshold, in an amount up to 50 percent of the contents of the Fund. The OSF was to be replenished with a third of overall excess revenues in years where total revenues exceeded budget estimates.5

32. In practice, however, higher oil revenues have continued to be used to compensate for shortfalls in other revenues, as was the case prior to the OSF. The revenue concept used to calculate transfers to the OSF includes total tax and oil revenue, such that oil windfall gains have often been used to compensate for tax revenue shortfalls. In 2000, of MEX$52 billion in higher-than-budgeted oil revenues, only MEX$9 billion was transferred to the OSF; in 2002, all excess oil revenues amounting to MEX$14 billion were used to compensate for tax revenue shortfalls, and in 2003 it is currently projected that, of MEX$47 billion in windfall oil revenues, only MEX$3.5 billion will accrue to the OSF. Moreover, in 2002, the rules for drawing on the resources in the OSF were changed to apply to any fluctuations in oil revenue, not just volatility caused by oil price developments. Simultaneously, by means of a temporary decree, the fund was depleted to offset large revenue shortfalls. It currently contains only MEX$75 million, or 0.001 percent of GDP— despite almost uniformly higher oil prices than envisaged in past budgets (Figure 4).

Figure 4:Evolution of the Oil Stabilization Fund

33. The adjustor on the amount of excess revenues to be saved has also been reduced incrementally. While only 1 percent of excess tax revenues over the first nine months of the year could be spent originally, in 2000 up to 0.13 percent of GDP in excess revenue could be spent (equivalent to about 1½ percent of budgeted tax revenues). From 2001, a third of excess revenue could be spent on investment; in 2003, the extra spending ceiling has been increased to 50 percent of excess revenue, and resources are to be turned over to the federal states. At the same time, shortfalls in tax revenue—if not offset by higher oil revenue—still have to be offset through cuts in expenditures.

34. Overall, these modifications, along with discretion in the implementation of the adjustors, have limited the deficit-reducing effect of the adjustors during economic upturns, while maintaining procyclical spending cuts during downturns. After depletion of the OSF in 2002, the regime in effect transmits any revenue volatility vis-à-vis the original budget (in part caused by oil price fluctuations) to spending programs, albeit to a lesser extent during periods of higher revenues. In other words, while the adjustors help to enforce the traditional deficit target, they leave little room for the limited automatic stabilizers that do exist to come into play, and reinforce the procyclicality of public finances, in particular during economic downturns.

35. In practice, the adjustors have been successful in reducing spending when revenue has fallen short, but failed to reign in expenditure during the boom year of 2000 (Table 3). While the adjustors were applied in 2000, in the sense that part of the additional revenue was saved and part was transferred to the OSF, additional spending was authorized beyond that implied by the adjustors, and a large reserve for potential tax reimbursements was created.6 In contrast, in 1998, 2001 and 2002, the government implemented expenditure cuts as revenues were below budget. In 1999, higher oil revenues and lower interest costs were to some extent offset by lower tax revenue, while the deficit still remained below target.

Table 3.Mexico: Budget Execution 1998-2002(In billions of pesos)
Budget 1/
Total revenueOil revenueOil price (US$)Total spendingTraditional balance
199880330015.5851(47)
19999502979.31,013(58)
20001,12437516.01,177(53)
20011,30241318.01,343(40)
20021,40539615.51,445(40)
Outturn 1/
Total revenueOil revenueOil price (US$)Total spendingTraditional balance
199878325310.2831(48)
199995631115.71,009(52)
20001,18842724.81,248(61)
20011,27138718.61,312(42)
2002 2/1,37740921.61,411(38)
Difference 1/
Total revenueOil revenueOil price (US$)Total spendingTraditional balance
1998(20)(47)-5.4(21)(1)
19997146.5(3)6
200063528.871(8)
2001(31)(26)0.6(32)(2)
2002 2/(28)146.1(34)3

Total spending and revenue numbers may not add up to the deficit shown above due to extrabudgetary operations.

Excluding Banrural, an operation for which higher spending was authorized by congress.

Total spending and revenue numbers may not add up to the deficit shown above due to extrabudgetary operations.

Excluding Banrural, an operation for which higher spending was authorized by congress.

36. Despite the adjustors’ relative effectiveness in meeting targets for the narrow deficit, they have no impact on fiscal consolidation. They are not legally binding, and are up for renewal with each annual budget. This has permitted congress to allocate a rising share of excess revenues to investment spending, instead of using windfall gains for debt reduction. Most privatization receipts are excluded. Also, public enterprises can spend higher revenues provided they meet their primary balance target, although some additional mechanisms are in place to reign in excessive spending in this sector. More generally, the adjustors are a limited tool that works only on the margin – or in relation to the budget that has been adopted in the first place, which in itself lacks a medium-term context; in addition, it focuses on the traditional, narrow deficit, as opposed to the augmented deficit, or net public sector borrowing requirement (PSBR).

D. The PRONAFIDE Framework

37. Each new administration in Mexico has developed a National Development Plan (Plan Nacional de Desarollo), backed up by sectoral plans, and a financing framework, PRONAFIDE. PRONAFIDE’s deficit and debt targets are not legally binding, have no formal connection to the annual budget process, and have not traditionally been updated during the life of the plan—all of which tend to limit PRONAFIDE’s relevance for annual budget discussions. For example, the Zedillo government’s traditional balance targets included in the 1997–2000 PRONAFIDE were missed by increasingly wide margins over the life of the plan, despite higher average growth rates and higher non-oil revenue than envisaged (Figure 5).

Figure 5:1997-2000 PRONAFIDE Traditional Budget Deficit Targets Versus Outturns

(In percent of GDP)

38. The current 2002-2006 PRONAFIDE targets for the augmented deficit and debt, published in 2002, are also at risk unless further consolidation measures are implemented. PRONAFIDE aimed at substantial fiscal consolidation and debt reduction between 2002 and 2006. The 2002 fiscal outturn and the 2003 budget objective for the “traditional” balance were in line with PRONAFIDE, excluding the restructuring of the rural development bank BANRURAL, and the objective for the PSBR was more than achieved in 2002. But the PSBR envisaged in the 2003 budget, as well as the outturn for the debt levels in the first quarter, exceed targets, and the gap would widen further unless new measures are introduced (Figure 6).7 The “no policy change” scenario shown assumes no tax reform, and primary spending remaining constant as a share of GDP at its 2003 level.8 Even adhering to this path would require considerable effort given existing spending pressures, as well as other risks to the outlook, such as the cost of a potential pension reform.

Figure 6:PRONAFIDE PSBR Deficit Targets and Projections under “No Policy Change”

(In percent of GDP)

Source: SHCP and staff estimates.

1/ PRONAFIDE midpoint between the two scenarios as shown in IMF Country Report No. 02/237.

39. Moreover, even if PRONAFIDE targets for the augmented balance from 2004 onwards were met, debt levels would remain above the path originally envisaged. Due to a combination of slow GDP growth in 2002-2003 and a rise in the peso value of external debt caused by currency depreciation, the ratio of augmented debt to GDP would remain above the PRONAFIDE path (Figure 7).

Figure 7:Gross Augmented Public Sector Debt

(In percent of GDP)

Source: SHCP and staff estimates.

E. A Fiscal Framework for Debt Reduction

40. Fiscal consolidation in many countries has been accompanied by institutional reform, notably by enshrining fiscal responsibility principles in policy frameworks. In Latin America, several countries have introduced fiscal responsibility laws, with varying degrees of success (Brazil, Argentina, Ecuador, Colombia and Peru), and Chile has adopted a structural balance rule. There is some evidence that fiscal rules have helped to fend off political-economy pressures, especially by imposing hard budget constraints on subnational governments. Sweden and the Netherlands constitute impressive examples of fiscal consolidation achieved in the context of rules-based frameworks.

41. Nevertheless, fiscal rules are not a panacea, and cannot substitute for true political commitment to structural reforms fostering long-term, sustainable spending adjustments. The failure of the Gramm-Rudman-Hollings framework to reign in the U.S. fiscal deficit in the late 1980s and early 1990s demonstrates the difficulty of achieving fixed targets absent political agreement on indispensable structural policy changes and without credible enforcement mechanisms (Figure 8). The U.S. experience in the 1990s with fiscal consolidation under the Budget Enforcement Act (BEA) in the context of greater political commitment was substantially more successful (Kell, 2002). Thus, fiscal rules can generally only strengthen policies that are already credible, as opposed to in themselves ensuring full credibility. Also, a simple, clear definition and complete transparency in fiscal accounting are key to ensuring the success of any rule.

Figure 8:The U.S. Deficit Compared with the Gramm-Rudman-Hollings Targets

(In billions of U.S. dollars)

Source: CBO (2003)

1/ Revised targets contained in the Balance Budget and Emergency Control Reaffirmation Act of 1987.

42. A framework for promoting further fiscal consolidation in Mexico could consist of two elements: fiscal responsibility principles to limit deficits over the medium term, combined with up-front tax reform measures. Tax reform to raise significantly non-oil revenues would enhance the authorities’ ability to run primary surpluses and reduce public sector liabilities without undue expenditure compression. However, broad political support for fiscal consolidation may be difficult to establish—not least since the narrow deficit usually under public scrutiny and the focus of budget discussions has diminished to 0.5 percent of GDP. In this context, additional safeguards in the form of fiscal responsibility principles could help to constrain spending pressures in the event of a tax reform to ensure further reductions in deficit and debt.

43. In 2001, the Mexican authorities presented a budget reform proposal to Congress. It included a constitutional amendment that would require that: the budget be set in a medium-term context, and be balanced over a four-year cycle; all excess revenues be saved; and expenditures adjust in the event of revenue shortfalls. Prospects for passage of the reform bill are currently uncertain. Also, it would apply only to the traditional deficit, as opposed to the PSBR.

44. Placing annual budgets in a medium-term framework with debt objectives and expenditure growth ceilings could promote greater discipline in budget discussions. At the same time, the risk of some activities being shifted outside of the traditional budget definition could be minimized by framing any fiscal responsibility principle in terms of broad measures of the deficit. Moreover, transmission of oil price volatility to spending programs could be avoided by focusing on the non-oil balance. Such a framework could also provide room for counter-cyclical policy once fiscal credibility has been fully established.9 A further refinement would be to incorporate borrowing by subnational governments in debt targets. While such borrowing has been stable and low in recent years at just under 2 percent of GDP, including it in the fiscal responsibility framework would help reduce its potential to become a significant contingent liability, as has occurred at times in the past. As recently as in 1996– 97, all states needed to be bailed out by the federal government following the 1995 tequila crisis (Giugale and others, 2000).10

Medium-term debt targets

45. Medium-term debt targets can help anchor a fiscal consolidation path. In principle, such a framework would allow temporary deviations, but fiscal slippages would trigger compensation over time for to safeguard the ultimate objective of fiscal consolidation: debt reduction. In practice, difficulties are manifold, however, and include determination of an optimal target level of public debt, and an optimal consolidation path; agreement on underlying macroeconomic assumptions; and identification of a credible mechanism to distinguish between temporary exogenous shocks—such as a temporary exchange rate depreciation, which would not require a fiscal response—and permanent changes that require fiscal adjustment. Also, an issue currently under discussion is the treatment of debt created to finance productive infrastructure investment.

46. Few countries implement explicit debt targets, and their experience has been mixed. For example, the Maastricht criteria for EMU membership included a quantitative target of 60 percent of GDP on public debt. New Zealand has endorsed the medium-term objective of keeping debt below 30 percent of GDP. Since 1998, Canada has committed to keeping the debt-to-GDP ratio on a downward track, with the objective of reducing central government debt to below 40 percent of GDP. Prior to the 2001 crisis, Argentina’s Fiscal Responsibility Law established limits on debt and expenditure growth (Craig and Manoel, 2002).

47. Various approaches have been developed to identify sustainable debt levels, which can help orient policy-makers by setting an upper limit to the target range. No specific level is universally applicable, of course. The Maastricht criterion of a maximum debt level of 60 percent of GDP is an example. While this might be a suitable ceiling for an industrial country with strong institutions for macroeconomic management, lower debt-to-GDP ratios of around 50 percent have been associated with financing problems in many emerging-market countries in the past (Reinhart and others, 2003; WEO, 2003). One approach is to establish a benchmark debt level by calculating the present discounted value of a country’s expected future primary surpluses based on past performance (WEO, 2003). The authors find that most emerging markets actually borrow 2½ times more than their fiscal track record would safely permit. Moreover, taking into account the volatility and uncertainty surrounding future revenues in the face of shocks to potential growth or commodity prices, the sustainable level of debt in countries vulnerable to such shocks would be even lower.

48. In the case of Mexico, the 2002 PRONAFIDE implicitly acknowledged the desirability of debt reduction by envisaging a 5 percent reduction in net debt between 2002 and 2006 from the initial level of 43 percent of GDP. The ultimate target for the debt ratio was not identified, however. Factoring in fiscal risks in the form of Mexico’s low revenue base and dependence on oil revenues by including shocks to growth and the oil price would reduce the sustainable debt level. By the same token, the sustainable debt level would rise if a comprehensive tax reform were to expand the tax base.

49. After the introduction of an explicit debt objective, mechanisms—insulated from the political process—need to be developed to achieve the target. In the case of Mexico, greater non-oil tax revenue mobilization would make the fiscal framework more resilient to spending pressures and social spending needs. At the same time, as discussed above, institutional changes would be needed to ensure significant saving of windfall revenues “for rainy days” and further debt reduction.

Medium-term expenditure ceilings

50. Growth of public spending could be limited to take advantage of favorable economic conditions to reduce the debt burden. Over the past 5 years, real primary spending grew by 1.2 percent on average—which seems restrained relative to potential GDP, but high when contrasted with real average growth in non-oil tax revenue of only 0.8 percent.

51. Many countries have implemented medium-term expenditure rules. Examples in Europe include Sweden, the Netherlands, Finland, and the United Kingdom, while the BEA in the United States set caps on growth of discretionary spending. In Latin America, Brazil’s Fiscal Responsibility Law limits the percentage of central government and subnational spending on public sector employees and prevents the creation of permanent spending mandates without corresponding increases in permanent revenue or cuts in other permanent spending items. Peru has limited annual growth in primary spending to the inflation rate plus 2 percentage points, and Argentina’s Fiscal Responsibility Law limited increases in primary spending to real GDP growth—to be kept constant in nominal terms should growth turn negative. In Colombia, the Subnational Fiscal Responsibility Law limits expenditures by subnational entities.

52. Several issues need to be addressed in implementing medium-term spending rules. General issues include: the margin of error of medium-term macroeconomic projections, or how conservative growth and inflation projections should be; the right measure of flexibility on the margin; and the importance of safeguarding the quality of spending, or of avoiding undue compression of certain expenditure items, such as infrastructure investment or social spending, or accumulation of wage arrears. Some countries have excluded certain spending programs from their rules for this reason.

53. In Mexico there are difficulties with controlling some spending components, especially wage expenses by public enterprises (Figure 9). Also, employment at the local government level—and associated wage expenses—have increased over the past decade, whereas federal employment has been reduced (Gil Díaz, 2001). One approach to addressing this problem would be to cap overall spending growth, especially of “permanent spending mandates,” such as wage outlays, or non-oil tax revenues—similar to the provision in Brazil. Moreover, enterprises could be made to operate in a more transparent and commercial framework. Any such provisions, however, would require addressing complex public sector contracts and other rigidities in public sector wage setting.

Figure 9:Average Real Growth of Selected Expenditures 1997-2002

Structural balance target

54. Once credibility has been established fully, Mexico could adopt a structural balance target to support economic activity through countercyclical policies. The large share of oil revenue in total revenue—about a third—suggests an approach that takes into account fluctuations in oil prices as well as cyclical swings in activity, as used in Chile with copper prices. Technical complications with this approach include identifying the long-term level of oil prices, making cyclical adjustments, and the difficulty of choosing a target level. In practice, it may also prove difficult to establish sufficient credibility to avoid the recurrence of financing constraints during economic downturns, which may have driven counter-cyclical policy in the first place.

55. Oil price movements are, not surprisingly, closely related to total budgetary oil revenue (Figure 10). As discussed above, the current OSF rules offer only limited protection against fluctuations in the oil price and revenues vis-à-vis the budgeted price and revenues— and since its depletion in 2002, no unanticipated oil revenue shortfalls can be compensated for.

Figure 10:Oil Prices and Revenues in the 1990s

56. A rough calculation adjusting primary balances throughout the 1990s for cyclical factors and changes in oil prices shows that adjustments would have been most significant during the tequila crisis and the boom year of 2000 (Figure 11). More often than not, these two adjustments would be mutually offsetting (Figure 12), in line with the previous finding that oil revenues tend to move counter-cyclically.11 In contrast, in Chile the output gap and copper income components are positively correlated (Philips, 2001). If the counter-cyclical behavior of oil revenues in Mexico were to persist into the future, the importance of making explicit adjustments in the budget for the cycle and oil price fluctuations would be reduced.12

Figure 11:Actual and Structural Primary Balances in Mexico in the 1990s

(Percent of GDP)

Figure 12:Structural Adjustment Components

(Percent of GDP)

F. Conclusions

57. Mexico has made significant strides in improving its fiscal position since the mid-1990s. But existing fiscal mechanisms have weaknesses from the point of view of ensuring further consolidation. The budget adjustors and the OSF, for example, only operate in relation to the initial annual budget assumptions, which in turn are not directly linked to the medium-term path for broad measures of the deficit and debt laid out in PRONAFIDE. As a result, the existing framework is not, in itself, likely to lead to further reductions in broad measures of the public deficit and debt, as envisaged in PRONAFIDE.

58. Progress in achieving sustained debt reduction would likely be furthered by enhancements to the institutional framework. These could include: comprehensive tax reform; focusing on augmented debt and deficit measures ; setting annual budgets in medium-term context – as envisaged in the draft budget reform; and implementing a fiscal rule constraining public expenditure growth, anchored by explicit, medium-term debt targets. Once the level of public debt has been brought down, and credibility has been established, the adoption of a structural balance target would provide scope for automatic stabilizers to moderate cyclical fluctuations within a medium-term framework.

References

    AlesinaA. and RobertoPerotti1996“Budget Deficits and Budget Institutions,”NBER Working Paper #5556.

    AlesinaA.RicardoHausmannRudolfHommes and ErnestoStein1999“Budget Institutions and Fiscal Performance in Latin America,”NBER Working Paper #5586.

    • Search Google Scholar
    • Export Citation

    BraunMiguel and Lucianodi Gresia2003“Towards Effective Social Insurance in Latin Amercia: The Importance of Countercyclical Fiscal Policy,”Paper prepared for the Annual Meetings of the Inter-American Development Bank and Inter-American Investment Corporation in MilanItalyMarch.

    • Search Google Scholar
    • Export Citation

    CalderónCesarWilliamEasterly and LuisServén2002“Infrastructure Compression and Public Sector Solvency in Latin America,”Central Bank of Chile Working Paper #187October.

    • Search Google Scholar
    • Export Citation

    Congressional Budget Office2003“The Budget and Economic Outlook: Fiscal Years 2004–2013,”Washington, D.C.January.

    CordobaJuan-Carlos2001“Alternative Measures of the Fiscal Deficit and Fiscal Sustainability,” inMexico: Selected IssuesIMF Country Report 01/191October.

    • Search Google Scholar
    • Export Citation

    CraigJon and AlvaroManoel2002“Budget Management Systems in Selected Anglo-Saxon and Latin-American Countries: A Comparative Assessment,”Paper prepared for the Conference on Rules-Based Fiscal Policy in Emerging Market Economies in OaxacaMexicoFebruary.

    • Search Google Scholar
    • Export Citation

    GavinMichaelRicardoHausmannRobertoPerotti and ErnestoTalvi1996“Managing Fiscal Policy in Latin America and the Caribbean: Volatility, Procyclicality, and Limited Creditworthiness,”Inter-American Development Bank OCE Working Paper No. 326March.

    • Search Google Scholar
    • Export Citation

    GavinMichael and RobertoPerotti1997“Fiscal Policy in Latin America,”NBER Macroeconomics Annual 1997. Cambridge, Mass: MIT Press.

    • Search Google Scholar
    • Export Citation

    Gil DíazFrancisco2001Speech prepared for the XXXI. Reunión Nacional de Funcionarios Fiscales in Ixtapan de la SalNovember.

    GiugaleMarceloAdamKorobow and StevenWebb2000“A New Model for market-Based Regulation of Subnational Borrowing: The Mexican Approach,”World Bank Policy Research Working Paper No. 2370Washington, D.C..

    • Search Google Scholar
    • Export Citation

    HicksNorman and Q.Wodon2001“Social Protection for the Poor in Latin America,”CEPAL Review Vol. 0 No. 73.

    KellMichael2002“The Effectiveness of the Fiscal Rules Under the Budget Enforcement Act and Options for Reform,” inUnited States: Selected IssuesIMF Country Report 02/165August.

    • Search Google Scholar
    • Export Citation

    Lloyd-SherlockPeter2000“Failing the Needy: Public Social Spending in Latin America,”Journal of International Development vol. 12 no. 1.

    • Search Google Scholar
    • Export Citation

    PerryGuillermo2002“Can Fiscal Rules Help Reduce Macroeconomic Volatility in LAC?”Prepared for the IMF/World Bank Conference on Fiscal Rules and InstitutionsOaxaca, MexicoFebruary.

    • Search Google Scholar
    • Export Citation

    PhillipsSteven2001“Chile’s Structural Balance Target: Design and Implementation Issues,” inChile: Selected IssuesIMF Country Report 01/120July.

    • Search Google Scholar
    • Export Citation

    ReinhartCarmen M.Kenneth S.Rogoff and Miguel A.Savastano2003“Debt Intolerance,” forthcoming inBrookings Papers on Economic Activity.

    • Search Google Scholar
    • Export Citation

    SchinasiGary and MarkLutz1991“Fiscal Impulse,”IMF Working Paper 91/91Washington: International Monetary Fund

    TalviE. and CarlosVégh2000“Tax Base Variability and Procyclical Fiscal Policy,”NBER Working Paper 7499.

    WodonQuentinNormanHicksBernadetteRyan and GabrielGonzalez2000“Are Governments Pro-Poor but Short-Sighted? Targeted Social Spending for the Poor during Booms and Busts,”WashingtonWorld Bank, mimeo.

    • Search Google Scholar
    • Export Citation

    World Bank2001“Mexico: Fiscal Sustainability, Volume II: Background Papers,”WashingtonWorld Bank, mimeo.

    World Economic Outlook2002“Debt Crises: What’s Different About Latin America?”in Chapter II: Three Essays on How Financial Markets Affect Real ActivityApril.

    • Search Google Scholar
    • Export Citation

    World Economic Outlook2003“Public Debt in Emerging Markets: Is it too High?”in Chapter IIIOctober.

1

Prepared by E. Jenkner.

3

The primary impulse is calculated using the standard IMF method for assessing discretionary fiscal policy changes (Schinasi and Lutz, 1991; Cordoba, 2001). In the case of Mexico, oil export receipts and nonrecurring revenue are excluded. Potential GDP and the output gap are estimated using a Hodrick-Prescott filter.

4

Budget laws contain other provisions to meet the deficit target, for example, that additional spending by public enterprises will only be authorized if it does not threaten the deficit target.

5

Transfers from and to the OSF are excluded from the augmented balance (PSBR) definition, which excludes all nonrecurring revenues; the traditional balance includes transfers from the OSF as revenues, and transfers to the OSF as an expense.

6

In 2000, total excess revenues amounted to MEX$59 billion. Of this amount, MEX$10 billion was transferred to a reserve fund for tax re-imbursements; an additional MEX$17 billion went to higher nonprogrammable spending, which exceeded estimates despite significantly lower interest costs, mainly due to high one-off transfers in debtor support; and the remaining amount— MEX$32 billion—was divided in accordance with the adjustors laid out in the law (Table 2), implying higher programmable spending of MEX$9 billion, transfers to the OSF of MEX$9 billion, and a “lower deficit” of MEX$14 billion.

7

Refers to the “inertial” scenario in PRONAFIDE.

8

Net of higher infrastructure investment financed out of excess revenues in 2003.

9

As a first step in this direction, fiscal transparency will be enhanced by the planned publication of a revised public sector definition in line with the new GFS by end-June, and the possible incorporation of medium-term macro-fiscal projections in the economic guidelines underlying the annual budgets.

10

In 2000, a new regulatory framework for debt management by states and municipalities was put in place, “outlawing” federal bail-outs, and requiring local governments to collateralize borrowing and to obtain credit ratings.

11

Tax revenues were adjusted for divergence from potential GDP with an elasticity of 1. Total oil revenues were adjusted for each percentage point fluctuation of the actual oil price from the average oil price for 1990–2002 with an elasticity of 0.4. This simple elasticity was estimated using the logs of quarterly oil prices and real oil revenues, seasonally adjusted, during 1990–2002.

12

In this context, the historical counter-cyclicality of oil revenues is not likely coincidental, as higher world oil prices depress activity in the United States and thus Mexico.

Other Resources Citing This Publication