The Selected Issues paper describes the nexus between household wealth, saving, and consumption, and provides estimates for the medium-term path of household saving and consumption. The paper also discusses to what extent the credit market frictions are holding back Ireland's economic recovery. Under current macroeconomic assumptions, the savings rate is expected to decline. Households have rapidly accumulated debt during boom times, and incomes and asset values have declined severely during the crisis. The Executive Board welcomes the country’s efforts toward economic recovery.

Abstract

The Selected Issues paper describes the nexus between household wealth, saving, and consumption, and provides estimates for the medium-term path of household saving and consumption. The paper also discusses to what extent the credit market frictions are holding back Ireland's economic recovery. Under current macroeconomic assumptions, the savings rate is expected to decline. Households have rapidly accumulated debt during boom times, and incomes and asset values have declined severely during the crisis. The Executive Board welcomes the country’s efforts toward economic recovery.

III. Medium-Term Fiscal Consolidation in Ireland: Growth-Friendly, Targeted, Sustainable1

A. Introduction

1. The Irish authorities have implemented a significant fiscal consolidation since the onset of the crisis: by end-2012, the structural primary deficit will have narrowed by 8 percentage points of GDP from its end-2008 peak of over 10 percent. While this has helped arrest the crisis-induced deterioration in public finances, significant further consolidation—5 percent of GDP, as per the authorities’ Medium-Term Fiscal Statement—is needed over 2013–15 to achieve their 3 percent of GDP target in 2015, and set debt on a downward path. With major decisions regarding this consolidation expected in the coming months, it is useful to take a strategic view of the consolidation strategy adopted thus far, and consider issues guiding the choice of future consolidation measures, taking into account longer-term trends in Ireland’s revenue and expenditures, and comparisons with other advanced economies.

2. Any future consolidation effort must rank high on the twin objectives of efficiency and equity. With domestic demand still fragile and unemployment at an elevated level, it will be important to avoid spending cuts or tax hikes with high multipliers or adverse employment or investment incentives, while favoring reforms that enable more cost effective delivery of key service priorities and a broadening of the revenue base. At the same time, the consolidation must be spread equitably across income groups, generations and family types, and protect the most vulnerable. Although Ireland’s poverty indicators remain better than much of Europe, the crisis has worsened poverty indicators for under-65s and inequality is edging upward. In this context, measures could focus on better targeting the state’s universal supports and subsidies (including on the tax side) and ensuring intergenerational equity, while also reining in demographics-related spending pressures.

3. Against this backdrop, this paper analyzes the evolution of Ireland’s revenue and expenditure ratios over time and from a cross-country perspective, with a view to informing discussions around the planned consolidation mix (section B); examines the current structure of taxation in Ireland, covering income tax, property taxation and environmental taxation, to indentify scope for revenue base broadening (section C); and evaluates potential high quality savings in health, education and social protection spending, as well as in the public service pay and pensions bill (section D).

B. Revenue and Expenditure Trends and the Consolidation Mix

Analysis of Trends in Revenue and Expenditures to Date

4. The structure of Irish public finances has undergone more than one significant transformation since the 1980s. As shown in Figure 1, in the 1980s Ireland was not a low-tax/low-spend economy. However, the sharp expenditure-led consolidation in late 1980s helped usher in a decade of break-neck export-led Celtic Tiger growth from the early 1990s, which saw the size of the public sector and revenues fall as a share of GDP through 2000.

Figure 1.
Figure 1.

Ireland’s Revenue and Expenditures

(percent of GDP)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: Eurostat and IMF staff calculations.

5. A turning point was reached around the turn of the millennium, when public finances of a much-richer Ireland were redirected toward expanding public services and the until-then lean welfare state. This “catch-up” spending happened well into the crisis, with welfare and pension rates rising by 3 percent as late as 2009, even though Ireland had entered a recession by early 2008. Figure 2 shows that social welfare rates doubled (more than tripled in the case of the universal child benefit) over 2000–09, and are 74 percent (130 percent in case of child benefit) higher than the level that would have obtained if they had grown at the rate of per capita nominal GDP. The exchequer pay bill, led by health and education, rose 118 percent, combining a 35 percent increase in personnel and 61 percent surge in pay, the latter outstripping cumulative per capita GDP growth by 16 percent.

uA03fig01

Ireland - Primary spending rose 140 percent over 2000-08

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Sourrce: Eurostat
Figure 2.
Figure 2.

Sources of Expenditure Increase During the 2000s

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: Departments of Public Expenditure and Reform, and Social Protection.

6. The late 1990s/early 2000s were also a turning point for the income tax. As Figure 3 reveals, Ireland’s 1980s configuration of a narrow income tax base with high marginal rates had been considerably reformed by the late 1990s. In the 2000s, policy focus shifted toward returning to the old-narrow tax bases, but unlike in the past, further cutting statutory rates. Thus, by 2009, the already high entry point for the income tax (at 25 percent of per capita GDP in 1998) had risen to over 50 percent of per capita GDP, while marginal rates had fallen to 20/41 percent, from 26/48 percent in 1998.

Figure 3.
Figure 3.

Income Tax Bands and Statutory Rates: 1985-2012

(bands in percent of per capita GDP; rates in percent)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: Department of Finance.

7. The very large deterioration in the underlying strength of the fiscal position implied by these spending increases and income tax cuts during the 2000s remained largely hidden under a flood of property-related revenues in the boom period of 2004–07. When the crisis erupted in 2008, boom revenues disappeared, output fell back sharply, and a structural primary deficit of over 10 percent of GDP emerged by end-2008, translating into underlying (i.e. excluding impact of direct banking support measures) general government deficits above 10 percent of GDP over 2009–10.

8. Figure 4 documents the dramatic rise of Ireland’s expenditure-to-GNP ratio from one of the lowest in the OECD in 2000, to one of the highest by 2011, while the revenue ratio has remained broadly unchanged. End-2011 current primary spending was 47 percent of GNP, 17 percentage points higher than in 2000: social benefits rose by 11.5 percentage points and public compensation by about 5 percentage points.2 Table 1 indicates that Ireland’s expenditures resemble those of English-speaking economies, as a share of GDP, but those of small European economies as a share of GNP. In other words, Ireland is either a low-tax/medium-spend economy (when scaling to GDP), or a medium-tax/high-spend economy (when scaling to GNP).3

Table 1.

Ireland’s Expenditure and Revenue Ratios vis-à-vis Comparator Groups

article image
Source: OECD and IMF staff estimates.Note: “English-Speaking economies” include U.K., U.S., Australia and New Zealand. “Small open European” economies include Austria, Belgium Denmark, Finland and Sweden (all these economies were sized between US$100–500 billion and had trade/GDP ratios above 80 percent over 2007–11); “Large European” economies include France, Germany and Italy (economies sized above US$2 trillion). OECD definitions for general government revenue and expenditure are used, which differ slightly from Eurostat definitions.
Figure 4.
Figure 4.

Expenditure-to-GDP, GNP ratios - Ireland Vs OECD

(2011, percent of GDP, GNP)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: OECD; and IMF staff estimates.

The Revenue-Expenditure Mix of Post-Crisis Consolidation

9. The crisis has prompted a sharp fiscal course correction in Ireland, with budgetary consolidation measures producing an 8 percent of GDP improvement in the structural primary balance over 2009–12. As Figure 5 shows, this large effort has been expenditure-led (two-thirds of total adjustment), combining a 14 percent cut in public wages; an 8 percent cut in welfare rates (except pensions); an almost 10 percent reduction in public service numbers from their 2008 peak; and savings in the non-pay current and capital budgets (17 and 63 percent, respectively, in nominal terms). Revenue contributions have included personal income tax (PIT) base broadening (10 percent reduction in income tax bands, introduction of universal social charge, elimination of Pay-Related Social Insurance (PRSI) reliefs and exemptions); higher taxes on capital and savings; and an increase in indirect taxes, most notably, the 2 point hike in the standard VAT rate to 23 percent in 2012. Progressive design and careful sequencing of this major consolidation, implemented during a deep economic slump, has helped preserve social cohesion, protect key public services and industrial peace, and maintain Ireland’s relatively strong poverty indicators within Europe.

Figure 5.
Figure 5.

Mix of Consolidation Measures, 2009–15

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: IMF staff estimates. SPB denotes structural primary balance ratio

10. With the overall deficit expected to still exceed 8 percent of GDP in 2012, the authorities are preparing for significant further consolidation over the medium term. The Medium-Term Fiscal Statement (November 2011) set out the parameters for a 5 percent of GDP consolidation over 2013-15 to deliver a deficit below 3 percent of GDP in 2015.4 The plan envisages a continuation of the expenditure-led approach (maintaining the two-thirds share), which can be justified given Ireland’s very high primary expenditure ratio in the OECD. Section D identifies significant scope for further expenditure savings, especially in health, education and social protection.

11. At the same time, it is important to recognize that the total expenditure effort envisaged is larger than that implied by the MTFS consolidation measures. As shown in the text table, and clear from Figure 6, the MTFS implies a reduction in primary expenditure-to-GDP ratio of 6.7 percentage points between 2012 and 2015, only half of which is to come from MTFS-announced measures; the remaining half is expected to arise from nominal freezes on welfare and pay rates.

Ireland: General Government Finances, 2007-15

(percent of GDP, excl. bank support costs)

article image
Source: IMF staff projections, based on MTFS consolidation path.
Figure 6.
Figure 6.

Projected Revenue and Expenditure Ratios Through 2015 1/

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: Eurostat and IMF staff estimates.1/ Figures are for general government.

12. Challenges to the achievement of this expenditure-led consolidation should not be understated. For example:

  • A weaker than expected nominal growth recovery could mean smaller expenditure-ratio reductions from nominal freezes in rates.

  • The demographic outlook will put significant upward pressure on pension, education and health spending over the medium to longer term. By 2020, Ireland’s population of over-65 year olds will rise by about one-third, and there will be about 15 percent more primary school age children than in 2012.

  • After several successive years of adjustment, few low-hanging fruit are left on the expenditure side.5 Indeed, the recent over-runs in health appear to have structural roots and unwinding them sustainably may require more fundamental reforms.

13. The alternative to an expenditure-led consolidation, of having a higher tax level to support high spending, akin to Northern European economies, would present questions such as:

(i) would revenue-based consolidation be less durable?

Mauro (ed.) et al (2011) find, for European countries over 1991–2007, that revenue-led consolidation plans were generally not common, but, where backed by concrete measures, these were generally implemented, and durably so.

(ii) would it be more contractionary for demand?

The influential work of Alesina and Ardagna (1998) and Alesina, Favero and Giavazzi (2012) suggests “yes”. In particular, the latter paper argues that revenue based-consolidations have been associated with long and severe recessions, while spending-led consolidations have been associated either with mild and short recessions or no recession at all. On the other hand, the large fiscal adjustment case studies covered in Horton et al (2004) and more recent work on fiscal multipliers (IMF 2012, see Appendix 1), find smaller growth costs of revenue-based consolidations in recessionary times because tax increases can induce lower private savings.

(iii) would high taxes/a larger public sector inevitably entail long-run growth costs and undermine Ireland’s FDI/trade-centered model?

Barro (1990) theorizes that diminishing returns from productive spending, and increasing costs of distortionary taxation, place an upper bound on the optimal size of the state. However, empirical studies have struggled to find a robust causal link from government size or aggregate revenue ratios to long-term growth and much depends on the mix of government – i.e. the productivity of spending and how distortionary are the taxes financing it. For instance, the small open economies in Northern Europe (Finland, Denmark and Sweden) have maintained relatively strong growth and competitiveness indicators despite their fairly large public sectors. However, expanding revenue for a larger public sector by raising already-elevated marginal tax rates (as in Ireland) would likely have long-run costs. Substantially higher effective rates on individuals and corporate could also impact Ireland’s attractiveness for high-skilled foreign professionals and foreign investors, clearly important considerations.

14. The foregoing suggests that no decisive conclusion on the appropriate revenue versus expenditure mix of fiscal consolidation can be reached. In that context, a pragmatic approach is to recognize that the chosen mix can be made more growth friendly by raising the productivity of spending and increasing reliance on less distortionary taxes. Given the still-fragile economy and high unemployment, it is vital that the choice of budget measures minimizes the drag on demand and job creation, while measures need to entail fair burden-sharing across income groups, generations and family-types, while effectively protecting the most vulnerable. This strategic approach, focusing on the quality (efficiency and equity) of measures, could involve base-broadening rather than rate hikes on the revenue side (section C), and better targeting of the state’s social supports and subsidies on the expenditure side, while reforming key government services (section D).

C. High-Quality Options for Revenue Base-Broadening

Relative to other OECD economies, Ireland has a combination of high personal and indirect tax rates and relatively narrow tax bases. Due to a very high entry point for income tax and employee PRSI, and despite elevated marginal rates for those around the average wage, Ireland’s effective PIT rates are quite low for persons earning up to 167 percent of average wage. Similarly, due to several lower-tier rates, the relatively-high 23 percent standard VAT rate currently applies to just 50 percent of consumption. This, together with the absence of a property tax, provides considerable room to raise revenues without raising the already high marginal tax rates, and while also avoiding higher rates on lower income workers that would undermine work incentives. There is also scope to expand the well-designed carbon tax to all fuel types, and to redesign vehicle taxes in a way that can provide higher revenues, while preserving incentives for environmental conservation.

15. Ireland has maintained a relatively low tax-to-GDP ratio over time (Figure 7). Current taxes on income and wealth, including social insurance contributions (i.e. direct taxes) are quite low, at around 18 percent of GDP – in fact, the lowest among advanced European economies. Indirect taxes (at 11.4 percent of GDP in 2011) were slightly below the OECD average, although pre-crisis the collection was close to 13 percent of GDP. At the same time, as shown in Figure 6, revenue as a share of GNP (arguably a better measure of the tax base) has traditionally been higher—varying between 40–50 percent of GDP—and is expected to approach 45 percent of GDP by 2015. This suggests that low revenue-to-GDP ratios may not automatically imply significant revenue-raising capacity, going forward.

Figure 7.
Figure 7.

Ireland Vs. OECD – Level and Structure of Revenues

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: OECD Revenue Statistics.

16. Table 2 sets out the shares of various taxes in total taxes (inclusive of social contributions), and reveals two interesting patterns. First, personal income taxes—at 47 percent of total taxes – are similar to the average share for English-Speaking economies. However, these economies compensate for their relatively low reliance on personal income taxes with higher shares of corporate income taxes and property taxes—about 23 percent; the figure for Ireland is less than 15 percent, with the shortfall primarily on account of property taxes. Instead, Ireland collects a relatively larger share from VAT, excises and non-fuel vehicle taxes.6 The following sections take a closer look at three taxes (income tax, property tax and environmental taxes), where there is scope for base-broadening and efficiency-preserving revenue raising.

Table 2.

Composition of General Government Revenues

Structure of Ireland’s Taxes Relative to Comparators (percent of total taxes)

article image
Source: OECD Revenue Statistics.

Personal income taxation7

17. Personal income taxes in Ireland comprise the income tax – which, in turn, is a combination of a core income tax (IT) and a universal social charge (USC) – and employee pay-related social insurance (PRSI).8 The IT accounts for about two-thirds of total personal income taxes, with the rest split roughly evenly between the USC and PRSI. The PIT structure and associated average and marginal taxes for a single PAYE taxpayer is summarized in Table 3.9

Table 3.

Ireland’s Personal Income Tax Structure

Structure of Personal Income Taxes: Rates, Thresholds and Average and Marginal Tax Rates

article image
Source: Department of Finance

18. As can be seen, Ireland has a fairly progressive personal income tax, with two characteristics that stand out and warrant further analysis:

(i) high marginal rates that kick in at fairly low income levels;

(ii) low average tax rates for most taxpayers, most notably for those earning between 67 and 167 percent of average wage.

High marginal rates

19. The first anomalous feature of the PIT system is the relatively low income level at which the top marginal rate kicks in. The higher income tax rate of 41 percent (and the corresponding top marginal rate of 52 percent, including the 4 percent USC and 7 percent employee PRSI) applies at €32,800, which is just above the average wage of €32,400. In this, Ireland is closer to the smaller European economies than the English-Speaking or Large European countries.

20. At end-2011, the level of the top marginal rate (52 percent) was high relative to the average for OECD and English-Speaking economies, but comparable to levels in the smaller European economies and the U.K., although in the latter, the rate will be reduced to 45 percent in 2013. Apart from generating efficiency concerns, the high top marginal rate in Ireland relative to the U.K. could also be problematic in terms of maintaining Ireland’s attractiveness as a location for high-earning professionals. On the other hand, it has to be noted that (i) location decisions are more a function of average, not marginal rates, which are not so out-of-line in Ireland for high-earners; (ii) the U.K. top rate will continue to kick in at a higher income level than in Ireland; and (iii) the distortionary effects of high top marginal rates are believed to be relatively small for high income-earners (Coady et al, 2012).

article image
Source: OECD.
Low average (or effective) rates

21. A comparison of average PIT rates over time below shows that Ireland’s current effective rates are in line with 2000 levels and that the sharp reduction in bands and credits through 2008 has since been clawed back. However, relative to most English-Speaking economies, and certainly the OECD average, the effective rates for an average-wage or below-average-wage single PAYE earner (especially if they are married with children) are quite low.10 Low average PIT rates have the advantage of reducing disincentives to take up work but, as noted by OECD (2011) and discussed in section III, benefits, not taxes, are the main drivers of such disincentives in Ireland.

Effective rates of personal income taxation

(including employee SSC, in percent)

article image
Source: Ireland Department of Finance and OECD (Taxing Wages, Comparative Tables, 2011).Note: Average annual wage (AW) in Ireland in 2012 = Euro 32,400.

22. Average PIT rates are low in Ireland for a number of reasons:

(i) The entry point for the core income tax regime is unusually high. The income level at which the IT rate of 20 percent kicks in for a regular single PAYE earner is €16,500.11 All income below this level is exempt from the income tax, irrespective of how high the total earnings of the PAYE earner are. The entry point threshold corresponds to 51 percent of the average annual wage (€32,400) which, by far, is the highest in the OECD: the next closest ratio is 27.6 percent (for Italy), while the average for both OECD and English-Speaking economies is 9 percent.12

(ii) The employee PRSI rate (at 4 percent) is modest and has an even higher exemption threshold (€18,304). Moreover, it comes with a generous initial “allowance”, i.e. deduction from income of the first €6,604 for the purposes of calculating the employee PRSI liability. Thus, the PRSI tax liability for someone earning €18,304 is 4% *(18304-6604) = €468, an average (or effective) tax rate of 2.6 percent.

(iii) The universal social charge (USC) applies to a broad base, taxing the entire income of any taxpayer earning above €10,036, but the USC tax rates (2, 4 and 7 percent) are quite low, so that the effective USC tax rate on incomes up to €16,500 (the income tax entry point) is just 2.9 percent. Moreover, special lower USC rates apply to medical card holders. Not only is this unusual (a health entitlement determining an income tax rate), the entitlement—at least in the case of the over-70s—is not subject to an effective means test, raising questions of fairness.

23. In addition to the standard regimes, there are a number of special tax reliefs that push up the entry point for income tax even higher for certain groups. For example, in addition to the two standard tax credits (the individual and PAYE tax credits), there are allowances (provided via credits or exemptions) for single parents, people who care for their children at home, people older than 65, rental expenses for older taxpayers, widows, handicapped and blind people. Similarly, lower USC rates apply to medical card holders, even when incomes are above those for non-medical card holders.13

24. Although intended to serve re-distributional purposes, such special reliefs are a poor instrument to achieve this result in a system where (i) tax credits are non-refundable (as they are in Ireland), i.e. they cannot result in a payment to the taxpayer in the case of tax liability being assessed as nil; and (ii) the sum of the two basic tax credits is so high: a low-income earner is already exempt from income tax, so that these special credits accrue mainly to middle or higher earners. Some of these special credits (and exemptions) have been scaled back in recent years, but scope for further tightening exists.

25. A reform strategy could seek to raise average PIT rates for taxpayers earning above 67 percent of average wage (or €21,708); increase the income level at which the top marginal rate kicks in; ensure better targeting of special income tax reliefs (including for USC); and smoothen out kinks in the tax schedule. The following is one way to achieve this:

  • a) Phase out the annual PAYE tax credit of €1650 between the minimum wage (€17,508) and the average wage (€32,400). This will increase the average and marginal income tax rates for persons earning between the minimum and average wage; raise the average tax for those earning above the average wage; and improve the targeting of special income tax reliefs.

  • b) The savings from (a) – which could be substantial – may be partly used to lower the income tax rate in the first bracket, or split it into two (e.g. 15 and 25 percent) so as to ensure that tax burdens do not rise for those earning below 67 percent of the average wage.

  • c) The income ceiling at which the top marginal rate kicks in could be increased somewhat to partly compensate those earning around the average wage, taking due regard of the scope that exists to ensure more equitable tax treatment of married couples vs. individual payers.

  • d) In addition, the PRSI could be better aligned with the income tax by (i) reducing the PRSI exemption threshold which, at €18,304 is 11 percent above the income tax entry point of €16,500; and (ii) phasing out the universal entitlement to an allowance on first €6,604 of income between the minimum wage and average wage, similar to what is proposed for the PAYE tax credit.

  • e) Finally, the interaction of the USC, income tax and PRSI could be reviewed to iron out large kinks in the average tax schedule at the USC and PRSI exemption thresholds.

Property taxation

26. Ireland currently maintains four types of property tax (the first is a transactions-based tax and the following three are recurrent):14

  • 2% stamp duty on non-residential transactions (with minimal exemptions) since 2012; and 1% (2%) stamp duty on residential transactions up to €1 million (on the balance above €1 million) since 2011.15 The combined collection from these stamp duties in 2010, i.e. before these lower rates were introduced, amounted to about €0.2 billion in 2010.

  • Commercial rates, which are collected by local governments, are based on the annual rental value of commercial premises (multiplied by a rate that is set by each local authority). The combined collection from these is around €2½ billion.

  • A non-principal private residence (NPPR) charge of €200 per NPPR, which yields a modest €65 million a year, although this is because the rate is low, not because of weak compliance. This charge was introduced in 2009.

  • A €100 household charge on principal private residence, introduced in Budget 2012, and initially expected to yield €160 million per year. Low compliance has meant the collection may fall short this year.16

27. The recent EC report on Taxation Trends in the European Union appears to suggest that, relative to Europe, Ireland has high taxes on property. However, comparisons with the OECD sample in Table 4, which includes English-Speaking economies with a tax structure more similar to Ireland’s (i.e. with low direct taxes), show a low level of property taxation in Ireland, especially for recurrent taxes on immovable property. For instance, in 2010, Ireland’s property tax take was 1.6 percent of GDP, compared with 1.8 percent of GDP for the OECD, and 3 percent of GDP for the four English-Speaking economies; the share of recurrent property taxation in total property taxes was 56.6 percent, well below the 83.3 percent in these economies.17

Table 4.

The Level and Structure of Property Taxation—Ireland vs. OECD

article image
Source: OECD Revenue Statistics

28. There are several arguments in favor of higher recurrent property taxation on immovable property (see Norregaard, forthcoming).18 First, they are a relatively stable source of revenue, which is important in small open economies with volatile tax bases such as Ireland. Second, they can promote efficient land use by imposing a “tax cost” on land ownership or use that to some degree may be independent from the actual use of the land (particularly if market price valuation is applied). Third, they can neutralize other distortions, such as more favorable income tax treatment of owner-occupied housing – e.g., due to mortgage interest reliefs – although this may not be as relevant in Ireland, as these reliefs are being phased out, and will be eliminated by 2017. Fourth, and perhaps most important, is the general acceptance of recurrent property taxes on immovable property as the least distortionary form of taxation in terms of reducing long-run GDP per capita, followed by taxes on consumption (and transactions), personal income and corporate income (OECD 2010).

29. The property tax was found to be regressive in its incidence by the earliest studies on the topic.19 However, more recent analysis finds the incidence of a property tax to be primarily on land and capital which, because they are owned predominantly by higher-earning individuals, implies progressivity. This view sees a property tax as a user charge for local public services (or their capitalized value, if market values are used), and thus as fair. Another aspect of fairness relates to the apportionment of property taxes between central or local governments. Because property values, in part, reflect services supplied by local governments, it is reasonable that they be allocated primarily to finance local activities.

30. The Irish authorities plan to introduce in 2013 a value-based property tax on principal private residences. Key design and implementation issues are:

  • Rate level: There would be little point in introducing the tax at a level below 0.2 percent. Given the importance of this reform for the future stability of public finances, and its relative growth-friendliness, consideration could be given to setting a rate comparable to levels in English-Speaking economies; for example, the average rate across the various states in the United States is 1.38 percent.20 A tax rate around 0.5 percent could yield annual revenue of €1 billion.

  • Rate setting: The power to set the rate could lie with the central government, or local government, or could be a mix of the two (perhaps central government setting a base rate, with local governments allowed to add a small top-up). That said, the property tax base should be clearly defined in central legislation, without any local discretion.

  • Exemptions and waivers: These should be kept to a minimum, although some allowance (perhaps in the form of a deferral, rather than outright waiver) for distressed mortgages could make sense.

  • Register: It would be critical to assemble a unified register of properties with details on who owns which property so that the liable taxpayers can be linked with Revenue’s database on tax numbers.

  • Collection: The local government’s challenges with the household charge, which had a lower-than-expected compliance rate, tilts the case in favor of central collection. In that case, the pros and cons of a PAYE-based deduction approach would need careful consideration due to the perceived impact on labor incentives, and administration cost and complexities for businesses.21

  • Apportionment of proceeds: Although details on revenue apportionment between the center and local governments are ultimately a matter for government, it would seem fair that a significant portion of the revenue intended for local governments go directly to the respective locality, while a sufficient amount be retained for equalization purposes (i.e. to be distributed to less-less-well-off local administrations).

  • Property valuation: As there is no up-to-date cadastre of property values in Ireland, and in an illiquid market many properties would be difficult to value in any case, the authorities may have to rely on a self-assessment regime initially. Although this is not a common approach, it has been implemented with some success in Latin America (Bogota City). Once the system is up and running, computer-assisted mass appraisal (CAMA) systems could be used for property revaluations.

Environmental taxation

31. Ireland’s environmental taxation comprises two types of energy taxes: excise taxes on motor fuels and a carbon tax that operates as a top-up on motor fuel excises, and applies separately to non-transport fuels used in industry and natural gas used in homes; and two types of vehicle taxes (increasing in CO2 emissions/km): one-off registration charges and annual motor taxes. In 2010, revenues from these taxes amounted to about 2½ percent of GDP in Ireland (down from 2.8 percent in 2000), compared with a weighted average of 1.7 percent for OECD countries as a whole. As a share of total tax revenues, environmental taxes were about 9 percent in Ireland, compared with a weighted average of 5½ percent of revenue for the OECD economies (Figures 8 and 9).

32. In Ireland, energy (fuel) taxation accounted for about 60 percent of 2010 environmental tax revenues, while vehicle taxes for the remaining 40 percent. The still relatively small collections from the carbon tax (introduced in 2011 and raised in 2012) are not reflected in these figures. Most other OCED countries raise disproportionately more revenue from fuel taxes, and some (like Denmark, Estonia and Netherlands) raise a significant amount from excise taxes on electricity.

Figure 8.
Figure 8.

Revenues from Environmentally Related Taxation

(percent of GDP)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: OECD.
Figure 9.
Figure 9.

Revenues from Environmentally Related Taxation

(percent of total revenues)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: OECD

33. The main recommendations in relation to these three environmental taxes are:

  • Domestic carbon tax: the level of the domestic carbon tax for emissions outside of the EU cap-and-trade system seems reasonable, but the current price for ETS emissions is only about a third of this level. A possible reform (to equalize emissions prices across the economy) would be to bring ETS emissions under the carbon tax, and provide refunds to covered sources for ETS allowance purchases.

  • Vehicle taxes: in the short term, a possibility for promoting stable revenue from vehicle taxes while at the same time more effectively exploiting technological possibilities for reducing emissions would be to implement a uniform ad valorem tax on vehicle sales combined with a ‘feebate’. The latter provides rebates for vehicles with relatively low emissions intensity and imposes fees on vehicles with relatively high emissions intensity.

  • Fuel taxes: there appears to be some modest scope to raise the level of passenger fuel taxes from the perspective of addressing adverse side effects of vehicle use including congestion, pollution, and road accidents. However, a more effective way to reduce congestion over the longer term would be to partly transition away from fuel taxes to a system of mileage-based tolls for busy roads with toll rates progressively rising and falling during the course of the rush hour.

Other taxation

34. Finally, we discuss some stand-alone tax issues that are likely to be important ingredients in any overall tax reform program: private pensions, PRSI base broadening, and VAT base broadening.

(i) Taxation of private pensions

About half of taxpayers maintain private pensions in Ireland. The tax treatment of private pensions is EET: tax-exempt at the contributions and accumulation, but taxed at distribution. However, because contributions can be deducted for tax purposes at the “higher” rate of income tax (41 percent), but capital gains and pension income would, in many cases, be taxed at 20 percent (e.g. if retirement income for a pensioner couple is below €36,000 p.a.), the current system subsidizes the contribution and accumulation stages beyond the incentives inherent in an EET system. These subsidies are poorly targeted, with richer taxpayers (who contribute more toward private pensions) receiving a substantial share of the subsidies. There is debate on the best approach to reduce or better target these subsidies:

  • Approach 1: Move to standard-rating of pension contributions deduction. The November 2010 National Recovery Plan had envisaged a gradual reduction, over 2012–14, in the rate of tax deductibility of private pension contributions from 41 percent to 20 percent. This would have preserved the EET regime for those expecting annual pension incomes (for a couple) below €36,000, but would have introduced some upfront taxation for those expecting larger pensions. The number of pensioners impacted could be large (as many as 650,000) so that the consequences of a large behavioral response that drives down long-term savings, could be quite negative.

  • Approach 2: Cap the cumulative amount of tax-relief-benefitting contributions. It has been suggested that instead of Approach 1, a cap on cumulative contributions of €1.5 million be set, “equivalent” to an annual pension payout in retirement of about €60,000. About 30,000 high-income pensioners would likely be affected in this case.

Overall, it is not clear why a combination of the two suggested approaches is not possible. First, the rate of tax relief could be consolidated for everyone at around 30 percent. This would have a relatively small effect on incentives to save for higher income taxpayers (subject to 41 percent tax rate), but could significantly raise the incentive to build retirement savings for those on lower incomes (subject to 20 percent tax rate). At the same time, a cap on cumulative relief-benefiting contributions seems fair. Whether €1.5 million cap is appropriate or is too generous, would have to be determined. It would be administratively easier, nonetheless, to apply the cap to new contributions, rather than retroactively.

(ii) PRSI Base Broadening

At 4 percent, Ireland’s employee PRSI rate is low compared with the mean average marginal employee social security contributions rate for OECD economies of about 8 percent in 2009. Relative to English-Speaking economies, the Irish rate is more comparable: the average of the marginal rates in the United Kingdom and United States was about 5 percent (Australia and New Zealand do not charge social security contributions). Given the large and widening deficit in the social insurance fund, however, the level of employee PRSI collections may have to be raised over the medium-to-long term (unless the state pensions are cut greatly). Given the already-high top marginal tax rate and the low threshold at which it kicks in, PRSI rate increases may be counterproductive and base-broadening should be the preferred course. Some of this broadening has already been done through the removal of the employer PRSI exemption on private pension contributions.

On the employee PRSI side, Budget 2012 signaled the possibility of extending PRSI to unearned income (i.e. rental, dividend or interest income), which would be a progressive base broadening measure. One concern is that the move would render the PRSI more like a tax, inducing adverse labor market incentive effects, and less like a hypothecated charge linked to some public or social benefit provision. However, as FAD (2012) notes, this issue is less pertinent in countries where the contributions-benefits link is tenuous. Ireland appears to one of these countries, as (i) entitlement to the full contributory pension (€230/week) is relatively easy to obtain (notwithstanding the more recent tightening of eligibility requirements); and (ii) even those not entitled to contributory pensions can get a non-contributory pension that is only marginally lower (€219/week). Overall, therefore, the broadening of the PRSI base to include un-earned income would not be expected to have significant adverse labor market effects given the de facto tax status of the PRSI.

(iii) VAT Base Broadening

Although Ireland’s collections from VAT are on the high side, and its 23 percent tax rate at the 75th percentile level in advanced economies, scope may exist for base-broadening. This is because the 23 percent standard rate applies to only about half of consumption. Ireland presently has three lower-tier rates (13.5 percent, 9 percent and 0 percent), in addition to two agriculture-related rates around 5 percent. The 9 percent rate was introduced in May 2011 as part of the government’s Jobs Initiative aimed at the services sector, including tourism, and it is scheduled to expire at the end of 2013. That would be a good opportunity to review the appropriateness of all lower-tier rates including the “zero” rate, which covers essential items like food, medicines and children’s clothing etc., as it is not efficient to give untargeted subsidies to everyone (rich and poor alike) through the VAT system.

D. Options for More Targeted and Efficient Expenditure

Ireland provides several expensive universal supports and subsidies, which are difficult to justify under present budgetary circumstances. Better targeting of spending, including the child benefit, medical cards, the household benefits package, subsidies on college fees, and non-means tested state pensions can generate significant immediate savings and contain demographics-related pressures over the longer-term, while effectively protecting the poor. While recognizing the benefits of the Croke Park Agreement, continued monitoring of the adequacy of savings in the net pay and pensions bill, and of service provision, is required, given the relatively high level of the public sector paybill. Longer-run reforms in health and education to ensure more cost-effective delivery of clearly-identified service priorities would be an essential compliment to ensure durability of savings over the medium-term.

35. Given recent staff analyses of social welfare spending and public pay and pensions (Box 3 and Box 5, respectively, in the IMF staff reports for the 5th and 6th EFF Reviews for Ireland), this section will briefly look at the following: (i) a discussion of service outcomes vs. 2010 expenditure levels in health, education and social protection; (ii) a brief overview of the expenditure effort thus far and planned; and (iii) an analysis of where targeting can generate significant immediate savings, while containing demographics-related pressures; and (iv) identification of longer-term reform priorities, leveraging the discussion in (i).

(i) Public Expenditure versus Outcomes

36. An analysis of Ireland’s expenditure on health and education (which account for more than half of total government expenditure) reveals a mixed picture regarding effectiveness (Tables 5 and 6). Spending on health grew rapidly between 2000 and 2010 to second highest in the OECD, and is now outsized relative to outcomes, which are mostly near the OECD average. Education spending (as a share of GDP), which was well below the OECD average in 2000, is now also slightly above it, while conclusions on outcomes are similar. Of the almost 5 percentage point of GDP increase over 2000–10 in health and education spending combined, four-fifths occurred in compensation to employees in these sectors. These trends suggest significant scope for efficiency savings, especially in health.22

Table 5.

Public Expenditure on Health and Education

(2011, percent of GDP)

article image
Source: OECD.
Table 6.

Selected Outcomes in Health and Education: Ranks in the OECD

article image

Rankings are organized such that a higher ranking (closer to 1) means a better outcome; hence a lower mortality rate and a higher life expectancy

out of 34 OECD countries

Source: www.oecd.org/health/healthpoliciesanddata/49105858.pdf
article image

out of 34 OECD countries except where otherwise indicated

Source: www.oecd.org/education/highereducationandadultlearning/48631582.pdf

37. While the benefits of the Croke Park agreement for industrial peace and efficiency-enhancing public sector reforms have to be recognized, Ireland’s public sector paybill still remains high. Ireland’s compensation of public employees (as a share of GDP) is only slightly above the 11 percent average European average level, but as a share of GNP, it is 3 percentage points higher.23 This of course, reflects the impact of the 35 percent increase in personnel numbers and the 61 percent surge in average pay rates over 2000–08, which have only partly been unwound: numbers have fallen by 9 percent, and pay rates by 14 percent (with some of the savings offset by rising public service pensioner numbers). Although cross-country comparisons of public wage premia are complicated by definitional issues, the simple metric of average public pay/GNP per worker appears quite elevated for Ireland relative to advanced European standards.24 This is consistent with findings in recent OECD surveys on health and education, which document above average pay levels in education (esp. secondary school teachers), and health (nurses and consultants), which together account for three-fourths of the public service.

38. With regard to social benefits, as noted in Table 1 previously, the rise in spending was even more pronounced, at almost 9 percentage points of GDP (7 percentage points, in structural terms) over 2000–11.25 As shown in Figure 10, the high level of benefits provides strong income protection to those on welfare (third highest level of protection in the OECD). This protection is most visible for those who lost employment during the crisis, but also to pensioners, whose poverty rates declined significantly over this period due to unchanged nominal pension rates and even rising real pension benefits since 2008.26

uA03fig02

Average public pay, as share of GDP per worker (2010) 1/

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

1/ Calculated as “compensation of general government employees-to-GDP” (Eurostat) × “total employment-to-general government employment” (OECD).
Figure 10.
Figure 10.

Net income level of person on social benefits in % of median household income (2010)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: OECD Tax-Benefit Model (update 3/31/12).

(ii) Overview of Expenditure Effort Thus Far and Planned

39. The Irish authorities adopted a range of expenditure measures since the crisis, totaling about 8 percent of GDP. These can be grouped into four broad categories: capital (25 percent of total effort), pay (25 percent), social welfare (17 percent), and other non-pay current spending (33 percent).27 The chronology of measures was as follows:

  • Late 2008/early 2009: “low-hanging” cuts in capital and program spending; hiring freeze

  • 2009–10: progressive wage cuts (14 percent); voluntary numbers reductions (ongoing)

  • 2010–11: welfare cuts of 8 percent (2010–11), esp. for families, but protecting elderly; 4 percent cut in public service pensions (2011);

  • 2012: Further capital cuts; higher user fees; some service reductions; many welfare measures, including changes to eligibility criteria (impact on many small groups); focus on non-core pay savings (overtime, sick pay, allowances)

40. For the period 2013–15, a further 3 percent of GDP in current expenditure measures (out of a total of 5 percent of GDP consolidation effort) is envisaged and most remains to be specified. Achieving this consolidation could be difficult for several reasons:

  • There are significant demographics-related spending pressures in education, health and pensions: the number of school-going children will be (15 percent) higher between end-2012 and 2020, and the number of over-65s one-third higher over the same period. In particular, the projected increase (3.5 percent of GDP) in Ireland’s pension expenditure between 2010 and 2050 is the fourth highest in the OECD (IMF, 2011)28

  • If the economic recovery were weaker-than-envisaged (and/or unemployment higher-than-expected), the implicit saving—in terms of reduction in expenditure-to-GDP ratio—from nominal wage and welfare freezes would be smaller. Indeed, welfare payments and the net public pay and pension bill have, thus far, fallen little as a share of GNP.

  • Some pockets of implementation shortfalls have begun to emerge, revealing structural roots which will take deep reforms to address (such as health system reforms).

  • Risks to services: as public service numbers are reduced and low-hanging fruit plucked, the difficulty of avoiding an adverse impact on public services and revenue collection capacity increases.

uA03fig03

Increase in Pension Expenditure over 2010-50

(percent of GDP)

Citation: IMF Staff Country Reports 2012, 265; 10.5089/9781475510461.002.A003

Source: IMF (2011)

(iii) Scope for Immediate Savings from Better Targeting

41. Against this challenging backdrop, the authorities must design a medium-term expenditure strategy that is durable, distributes the adjustment burden fairly across income groups (with the most vulnerable being effectively protected), generations (old vs. young), and household types (single vs. families), including due consideration of pre-crisis spending increases and relative adjustments since then. Table 6 lists key spending items where targeting can deliver substantial, durable and progressive savings. Among these are three universal supports and subsidies (child benefit, household benefits package and subsidy on student fees) which amount to 2½ percent of GDP. Although medical cards are means-tested for everyone, almost 95 percent of persons over 70 years old qualify, given a significantly more generous means test for this group. Overall, staff estimates that realistic percentage reductions in these spending areas can yield permanent annual savings of about around 2 percent of GDP. Critically, these savings will contribute directly to containing the impact of demographics-related changes to the spending profile. Table 8 details further the rationale for better targeting of spending in these areas.

Table 7.

Selected Expenditure Items Offering Scope for Targeting/Savings

article image
Source: Revised Estimates Volume (2012), and IMF staff estimates.
Table 8.

Rationale for Targeting Reforms in Selected Expenditure Items

article image
article image
Sources: Factual information underpinning staff analysis was provided by the Departments of Public Expenditure and Reform, Health and Education. Other sources consulted include: OECD Health at a Glance (2011) and OECD Education at a Glance (2011).

(iv) Need for Longer-term Reforms

42. In parallel to achieving near-term targeted savings, reforms of key public services are needed to underpin savings in the medium term. As noted earlier, despite spending substantially more than the OECD average on health, Ireland’s performance indicators are only at/or marginally above OECD average, with similar results in education while spending is modestly above OECD average. Hence, there is need for deeper reforms in these areas to identify service priorities and deliver them more efficiently, including by fully utilizing the flexibility provided by the Croke Park agreement. For example, new working models to minimize premium and overtime payments, and a substantially greater use of primary vs. hospital care and generic vs. branded drugs, can significantly reduce the public cost of healthcare while preserving outcomes.29 To better focus health spending on programs and outcomes, it will be important to extend performance budgeting to the sector, and implement governance reforms that enable cost priorities set at the center to be reflected in decisions at the local and hospital level.30

43. Similarly, a new funding model for higher education, that (i) better takes into account emerging skills priorities/shortages, with some linkage of college fees to cost/earnings potential of courses (and supported by affordable loans and grants for poor students), and (ii) strikes the appropriate balance between degree and vocational education (where the latter’s share is just 10 percent, one-third the OECD average), can deliver broad access to high-quality education that underpins Ireland’s competitiveness without additional public investment. While savings from these health and education reforms can take time to realize, they are important to help ensure the consolidation can be sustained by enabling the growing needs for these services to be met at manageable cost.

1

This paper was prepared by S. M. Ali Abbas with substantial inputs from Ruud de Mooij, John Norregaard, Ian Parry, Baoping Shang and Mauricio Soto (all FAD). Asad Zaman and Ari Binder provided outstanding research assistance. The author is also grateful to staff in Ireland’s Departments of Finance, Public Expenditure and Reform, Health and Education, and Irish Revenue, for providing data and for helpful suggestions.

2

The bulk of this increase in the primary current expenditure-to-GNP ratio occurred between 2007 and 2011, given the crisis-induced peak-to-trough collapse of 18 (24) percent in nominal GDP (GNP). However, absent a commensurate expected boom in output, the very high spending ratios today are largely structural.

3

Given the unusually large and expanding wedge between GNP and GDP in Ireland, it is instructive to scale fiscal variables to both measures of economic activity. Ratios to GNP may, in fact, be more relevant, insofar as GNP is more closely linked to tax revenue.

4

The authorities expect to update this statement in October 2012 in preparation for Budget 2013.

5

Evidence from past European consolidations, especially the recent Latvian experience, suggests that expenditure reductions become progressively more difficult in later phases of consolidation, or when key services begin to be affected. For evidence on past European consolidations, see Abbas et al (2011). For a summary view on the Latvian case, see Blanchard (2012).

6

Indeed, taxes on production and imports (i.e. indirect taxes) are comparable to EU levels (in percent of GDP), so that as a share of total taxes, the reliance on non-distributive taxes is quite high.

7

Issues related to pension tax reliefs and PRSI base broadening to unearned income are discussed under “Other”, toward the end of this section.

8

The Irish welfare system does not differentiate significantly between social insurance and social assistance, or between contributory and non-contributory state pensions. Accordingly, PRSI contributions do not bear a strong link to welfare benefits, so that it is acceptable to combine (employee) PRSI with income tax and USC when looking at personal income taxation in Ireland.

9

Note that Table 3 uses the term “exemption threshold” to connote the level of income that, if one earns below it, no tax is incurred. In this sense, both the USC and PRSI have exemption thresholds: i.e. those earning below €10,036 pay no USC, and those earning below €18,304 pay no employee PRSI. However, those earning above these levels pay the said tax on all income, including the income below the threshold, implying a jump in the effective tax schedule around the threshold. By contrast, the “entry point” for income tax refers to the amount of income on which there is no tax payable at all. Thus, a person earning €16,499 pays zero income tax, a person earning €16,500 pays 20 percent of €1 (= €16,500-16, 499) or €0.2 in income tax, while a person earning €30,000 pays 20 percent of €13,501 (= €30,000-16, 499) or €2,700 income tax (which is 9 percent of the gross income of €30,000).

10

The effective tax rates for self-employed individuals and couples are much higher, as they are not entitled to the PAYE tax credit. However, these taxpayers are entitled to deduct business expenses from income in their tax returns, so the actual tax take for the government would not be much higher.

11

The entry point for the income tax (€16,500) for a PAYE earner is determined by the combination of the basic tax credit of €1,650 and the PAYE credit (similar to an earned-income tax credit) of €1,650. For earnings below €16,500, the annual tax liability would be less than €3,300 given the (lower) 20 percent income tax rate, which would be covered by the sum of the personal and PAYE tax credits (each €1,650).

12

Note that €16,500 is only slightly lower than the minimum wage of €17,542: €8.65/hour*39 hours a week*52 weeks a year.

13

This is possible in the case of the over-70s who still have near-universal entitlement to the medical card. Separately, it is quite unusual for a health entitlement to determine the PIT tax rate applicable to a person.

14

Like most other OECD economies, Ireland does not have a net wealth tax.

15

These rates have been brought down significantly from the 7-8 percent prevailing before the crisis.

16

Ireland had a recurrent property tax on principal residence of 1.5 percent till 1997, when it was abolished. This tax was levied on the excess of the market value of all relevant residential properties of a person over a market value exemption limit and was payable provided the income of the household exceeded an income exemption limit.

17

The 2010 data suggests a high share of transactions taxes in Ireland (mainly stamp duties), but these have been brought down since then.

18

Taxing Immovable Property: Revenue Potential and Implementation Challenges”, IMF Staff Discussion Note.

19

These studies saw a property tax as a combination of a tax on mobile capital and a tax on immobile land (the former got shifted to renters, consumers, and labor, while the latter borne by landowners).

20

Based on U.S. property tax rates, as reported in State-by-State Property-Tax Rates.

21

Persons on pensions/welfare would presumably require such deduction to be implemented by the Department of Social Protection, which could undermine somewhat the perceived benefits of collection by a single central agency.

22

Based on data through 2008, Grigoli (2012) and Clements et al (2012) reach similar conclusions. OECD (2010) finds Ireland to be among the least efficient even within countries with similar health care systems.

23

See comparisons in a recent staff analysis of savings in the public wage bill (Box 5 in Ireland – Sixth Review Under the Extended Arrangement – IMF Staff Report).

24

A forthcoming study by the Central Statistics Office is expected to shed some light on the question of a public wage premium in Ireland.

25

For a recent staff analysis of social welfare spending, see Box 3 in Ireland – Fifth Review under the Extended Arrangement – IMF Staff Report.

26

Box 2 in the accompanying Article IV staff report shows that in 2010, Ireland had the second lowest relative at-risk of poverty gap in Europe, after Finland, while consistent poverty, at 6.2 percent, remains below 2006 levels.

27

These shares are based on announced measures (relative to the baseline of no policy change), but adjusted by staff for baseline realism.

28

The EC’s 2012 Ageing Report estimates the increase in gross pension expenditure between 2010 and 2060 at 4.1 percent of GDP, the 8th highest increase in Europe (despite the rise in the retirement age to 68 by 2028). Moreover, the benefit ratio (average pension to average wage) is expected to rise, making Ireland an outlier in Europe (along with the U.K.).

29

Generic drugs account for only one-fifth of all prescriptions, which compares poorly to four-fifth in neighboring United Kingdom.

30

The recent HSE Governance Bill should be helpful in this context.

Ireland: Selected Issues
Author: International Monetary Fund