Journal Issue
Share
Chapter

CHAPTER 2 Expenditure Reform

Author(s):
Benedict Clements, Juan Toro R., and Victoria Perry
Published Date:
October 2010
Share
  • ShareShare
Show Summary Details

Expenditure Reform: Key Principles

Expenditure reforms should be guided by two objectives:

  • Improving the efficiency of spending. Countries should seek to reduce the cost of producing existing public sector outputs. In addition, spending should be allocated to activities that provide the greatest marginal benefits to society as a whole; and
  • Ensuring equity. Growth without equity is less durable (Berg, Ostry, and Zettelmeyer, 2008; and Tanzi, Chu, and Gupta, 1999). Expenditure policy must reflect the need for both intra- and intergenerational equity. In view of demographic pressures, ensuring intergenerational equity will require altering the terms of social insurance in many countries. Greater targeting of social spending may also be necessary to ensure that the poor are protected as spending levels are reduced as a share of GDP.

Expenditure Structure and Trends

Cross-country differences in the size and composition of government spending are sizeable, reflecting differences in the level of development, role of the state, and spending efficiency. Expenditure is generally higher in advanced economies, reflecting more expansive social benefits (Tables 2.1a and 2.1b). There are also significant variations within the advanced economies, reflecting differences in demographic structure and sociopolitical preferences regarding the role of government. Outlays for the wage bill are higher in advanced than emerging economies. Capital expenditures are generally higher in emerging economies, but with wide variation across countries. Although high spending alone does not indicate inefficiency, several studies suggest that many countries could achieve similar levels of public services in education and health at a lower cost (Carcillo, Gunnarsson, and Verhoeven, 2007; and Afonso, Schuknecht, and Tanzi, 2005 and 2006).

Table 2.1a.Expenditure Structure: Advanced Economies, 2008
(In percent of GDP)(In percent of primary expenditure)
Primary expenditureCompensation of employeesSocial benefitsCapital spendingOtherCompensation of employeesSocial benefitsCapital spendingOther
Australia 1230.48.89.62.69.428.931.68.630.9
Austria 346.39.223.61.112.419.951.02.426.8
Belgium 346.212.123.31.79.126.250.43.719.7
Canada 436.111.67.51.415.632.120.73.943.3
Cyprus 339.814.112.13.010.635.430.47.526.6
Czech Rep. 441.87.618.25.011.018.243.512.026.3
Denmark 350.517.316.41.815.034.332.53.629.7
Finland 347.513.417.72.613.828.237.35.529.1
France 349.912.723.33.210.725.546.76.421.4
Germany 341.06.924.31.58.316.859.33.720.2
Greece 343.711.519.12.910.226.343.76.623.3
Hong Kong 1216.74.24.40.08.125.026.20.048.7
Iceland 354.414.66.14.529.226.811.28.353.7
Ireland 341.011.113.85.310.827.133.712.926.3
Israel 240.212.19.30.118.730.123.20.246.5
Italy 343.610.920.42.210.125.046.85.023.2
Japan 5633.56.217.73.66.018.452.910.818.0
Korea 5629.17.35.95.710.225.120.119.635.2
Luxembourg 337.47.118.13.68.619.048.49.623.0
Malta 341.714.613.32.511.335.031.96.027.1
Netherlands 343.89.120.23.511.020.846.18.025.1
New Zealand 6738.19.313.33.312.124.435.08.731.8
Norway 338.512.013.63.19.831.235.38.125.5
Portugal 343.012.919.92.28.030.046.35.118.6
Singapore 1214.64.23.11.55.828.921.510.339.4
Slovak Rep. 333.66.615.62.09.419.646.46.028.0
Slovenia 343.111.116.74.311.025.838.710.025.5
Spain 339.510.815.03.89.927.338.09.625.1
Sweden 351.314.918.23.314.929.035.56.429.0
Switzerland 1331.07.711.61.99.824.837.46.131.6
United Kingdom45.011.013.12.318.624.429.15.141.3
United States 436.110.212.91.012.028.235.72.933.3
Average38.19.615.32.111.125.339.65.729.3
Advanced G-2037.79.515.22.011.125.239.85.529.6
Sources: WEO; Eurostat; Government Finance Statistics (GFS); and OECD.

2007 data.

GFS.

Eurostat; capital spending proxied by “gross fixed capital formation.”

WEO.

2006 data.

OECD; capital spending proxied by “gross fixed capital formation.”

2005 data.

Sources: WEO; Eurostat; Government Finance Statistics (GFS); and OECD.

2007 data.

GFS.

Eurostat; capital spending proxied by “gross fixed capital formation.”

WEO.

2006 data.

OECD; capital spending proxied by “gross fixed capital formation.”

2005 data.

Table 2.1b.Expenditure Structure: Emerging Market Economies, 2008
(In percent of GDP)(In percent of primary expenditure)
Primary expenditureCompensation of employeesSocial benefitsCapital spendingOtherCompensation of employeesSocial benefitsCapital spendingOther
Argentina 1230.410.112.63.83.933.312.6
Brazil 1232.59.712.72.67.529.98.0
Bulgaria 336.59.012.05.79.824.732.915.626.8
Chile 1422.65.34.64.18.623.620.317.938.1
China 120.5
Colombia 123.05.67.55.14.924.232.521.921.3
Croatia 5640.29.915.54.210.624.638.610.426.4
Egypt 631.07.14.05.214.622.913.016.847.3
Hungary 345.111.518.72.812.125.541.56.226.8
India 1223.64.1
Indonesia 1718.64.60.85.77.524.930.5
Malaysia 1227.36.53.011.0
Mexico 1822.25.97.45.53.426.624.8
Nigeria 128.2
Pakistan 1217.41.9
Peru 117.35.12.33.86.029.713.322.234.8
Philippines 1213.15.03.12.42.738.118.1
Poland 341.110.016.24.610.324.339.411.225.1
Romania 337.610.211.25.610.627.129.814.928.2
Russia 134.07.69.46.510.522.327.619.231.0
Saudi Arabia 1829.910.62.27.49.735.624.6
South Africa 127.49.710.71.95.135.439.07.118.5
Turkey 128.46.910.53.87.224.337.113.325.3
Ukraine 646.510.619.55.710.722.741.912.323.1
Average25.87.910.74.77.726.922.716.821.0
Emerging G-2024.87.910.84.87.327.320.417.518.1
Sources: WEO; Eurostat; GFS; ILO; and IMF staff estimates.

WEO.

ILO Social Security Department database Global Extension of Social Security (GESS), accessible at http://www.socialsecurityextension.org/gimi/gess/ShowTheme.do?tid=1985.

Eurostat; capital spending proxied by “gross fixed capital formation.”

“Social benefits” include social security benefits only.

2007 data.

GFS.

Public pensions only.

IMF staff estimates.

Sources: WEO; Eurostat; GFS; ILO; and IMF staff estimates.

WEO.

ILO Social Security Department database Global Extension of Social Security (GESS), accessible at http://www.socialsecurityextension.org/gimi/gess/ShowTheme.do?tid=1985.

Eurostat; capital spending proxied by “gross fixed capital formation.”

“Social benefits” include social security benefits only.

2007 data.

GFS.

Public pensions only.

IMF staff estimates.

Age-related spending has been the main driver of current spending over the past two decades. Within the advanced countries, age-related outlays have risen since 1990 by roughly 2 percentage points of GDP (Figure 2.1). Increases have been especially large for pensions in Japan and Korea in the past decade, and for health spending in Korea (from a low level) and France, Greece, and Portugal. Demographics have been an important catalyst behind these increases, particularly in pensions, in the advanced economies. For health, technology4 and its interaction with an ageing population have been the key drivers behind rising spending. These trends are expected to continue in the coming years for both advanced and emerging economies (see discussion below on age-related spending).

Figure 2.1.Age-Related Expenditure Trends in the Advanced Economies

(In percent of GDP)

Sources: OECD; WEO; and IMF staff estimates.

Note: Countries included in the sample are Australia, Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States.

Reinforcing past trends, primary current expenditures, adjusted for the cycle, have risen further during the crisis. After the success of the 1990s in containing spending increases as a percent of GDP—owing to reductions in non-age-related outlays—primary current spending began drifting upward in the years prior to the crisis, but with wide variation across countries (Figures 2.1 and 2.2; and Figures A.1 and A.2). In some countries with moderate and high adjustment needs, current primary spending was already rising in 2000–07, including for wages (Italy, the United Kingdom, and the United States). Primary current spending also increased steadily in many emerging economies such as Argentina, Brazil, Mexico, and South Africa (Figure A.3). Since the crisis, current outlays have accounted for the bulk of the rise in spending in the advanced G-20, partly reflecting safety net spending. Of the increase in structural spending of about 2½–3 percentage points of GDP in the advanced G-20 economies between 2007 and 2010, about 1½ percentage point can be attributed to discretionary stimulus measures. In the emerging G-20, primary balances have weakened due to higher spending, especially for social benefits, the wage bill, and public investment.

Figure 2.2.Primary Expenditure Trends in the Advanced Economies

(In percent of GDP)

Sources: OECD; WEO; and IMF staff estimates.

Notes: Countries included in the sample are Australia, Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States; and the dashed line represents percent of potential GDP.

Expenditure Reform Strategy

In countries requiring fiscal consolidation, the adjustment on the spending side will have to go well beyond the expiration of the stimulus spending increases; more fundamental reforms are needed. Non-renewal of the stimulus spending would lower spending by 1½ percent of GDP in the advanced and emerging G-20 economies, only a fraction of the needed adjustment.

A two pillar strategy could anchor expenditure reform, guided by the following objectives (Ali Abbas and others, 2010):

  • Stabilize age-related spending relative to GDP. Given the major trend increase in these outlays, reducing this spending would be difficult. The goal should thus be to stabilize spending-to-GDP ratios, which will require significant structural reform.
  • Reduce non-age-related spending relative to GDP. A possible policy goal would be to stabilize aggregate non-age-related spending in real per capita terms so that the ratio of GDP drops as growth picks up. In the advanced economies, for example, freezing non-age-related spending in real per capita terms over the next 10 years—beyond the savings arising from the non-renewal of stimulus spending—could generate structural savings of about 3–3½ percentage points of GDP.5 Similar policies helped underpin some successful fiscal consolidations in the 1980s and 1990s, such as in Belgium (1983–89), Denmark (1982–86), Finland (1993–2000), Israel (1980–83), and Sweden (1993–2000).

To achieve these goals, medium-term expenditure reforms will need to improve the composition and efficiency of expenditure. The freeze in real spending is an overall policy goal, not a tool. Targeted structural reforms would be needed to achieve this goal. In both advanced and emerging economies, reforms in wages, subsidies, and transfers have been the most durable and conducive to economic growth.6 Staff analysis of the experience with large fiscal adjustments provides a similar picture, with cuts in the wage bill comprising about a quarter of the adjustment and social benefits and transfers accounting for almost a third (Table A.1). Containing age-related spending has also been an important element of the adjustment.7 A breakdown by functional classification shows that reductions in general public services, economic affairs, and defense spending have comprised an important element of adjustments among advanced economies in these episodes (Table A.2).

Better targeting of social welfare spending, including social benefits, could provide substantial fiscal savings. Social benefits are large—both in percent of GDP and as a share of spending—in many countries with high adjustment needs (Tables 2.1a and 2.1b).8 Much of this spending, however, is not well targeted. In the OECD, less than 10 percent of public social spending is means-tested (Adema and Ladaique, 2009). This partly reflects a high share of age-related, insurance-based outlays in social spending. Nonetheless, the effectiveness of cash transfers in reducing inequality varies considerably, even among countries with similar systems (ILO, 2010, OECD, 2008a). In the European Union, less than a third of non-age-related benefits are means-tested (Figure 2.3). This suggests substantial scope to reduce these outlays without sacrificing equity objectives. Improved targeting of tax benefits (including for employer-provided benefits) should also be explored, with due consideration of the implied increase in marginal tax rates (see also discussion in Chapter 3 on personal income taxes).

Figure 2.3.Targeting of Non-Age-Related Social Spending, 2007

(In percent of GDP)

Source: Eurostat.

Note: Non-age-related social spending reflects social benefits excluding old-age and health spending.

The design of unemployment benefits could be improved. As employment recovers, spending programs providing long-term assistance to the unemployed should be reexamined. The high long-term replacement rates of unemployment benefits in some countries, for example, can have high fiscal costs and adverse labor market effects (OECD, 2009b) (Figure 2.4). Staff estimates that in countries with above-average replacement rates, reducing them to the OECD average could yield savings of almost ½ percent of GDP. Efforts to tighten the duration and generosity of out-of-work benefits, with increased emphasis on in-work benefits as well as a tightening of eligibility for sickness and disability benefits, would minimize disincentives for labor force participation (Carcillo and Grubb, 2006). These policies will be all the more important in the context of declining labor supply due to ageing populations.

Figure 2.4.Generosity of Long-Term Unemployment Benefits

(In percent)

Source: OECD (2009b).

Note: Unweighted averages, for earnings levels of 67 percent and 100 percent of average wage and four family types. Includes cash incomes, income taxes, and social security contributions. Excludes social assistance or housing related benefits. For further details see www.oecd.org/els/social/workincentives.

Reversing recent increases in military expenditure could also yield savings. Reducing outlays in the advanced economies to pre-crisis levels could generate savings of about ½ percent of GDP (Figure 2.5). Returning spending to levels prevailing roughly a decade ago would yield 1 percent of GDP.

Figure 2.5.Military Expenditure in Advanced Economies, 1990-2010

Sources: Stockholm International Peace Research Institute (SIPRI) military expenditure database; and WEO.

Sizable savings are possible in spending on subsidies. Subsidies averaged about 1 percent of GDP in 2007 in OECD countries, and equaled or exceeded 2 percent of GDP in Austria, Belgium, Denmark, and Switzerland. This spending, including for agricultural subsidies—which are large in some countries—should be reexamined and replaced, where possible, with more targeted instruments to provide income support. In particular, priority should be given to phasing out energy subsidies, including for petroleum products. Tax-inclusive subsidies for these products, which also incorporate estimates of the needed taxation to offset externalities, are projected to reach 1 percent of global GDP in 2010 (Coady and others, 2010) (see also below). Advanced economies account for about a quarter of this total, and emerging countries over half.

Public spending on climate change is expected to increase, but this can be moderated by improving the efficiency of these outlays.9 Subsidies for renewable electricity and biofuels may have become excessive.10 Potentially more productive spending to address climate change includes programs for energy R&D and low-carbon or climate-resilient infrastructure. While additional expenditures (in some cases substantial) are needed to address climate concerns in advanced and emerging countries, the primary focus of climate policies should be to reduce emissions through the appropriate carbon pricing (see below). Even beyond countries’ domestic climate policies, increased public expenditures from advanced countries will be needed to help meet commitments to support adaptation and mitigation in developing countries (pledged to reach $100 billion by 2020).

Expenditure reviews could help guide the design of country-specific strategies. These reviews, which have played a key role in expenditure reform in several advanced economies, can provide valuable input to guide long-term reform by addressing fundamental questions on the role of government and the cost effectiveness of different policy interventions (IMF, 2008b; and Kelly, 2007). These reviews should also identify expenditure inefficiencies and be integrated with performance-based budgeting.

Coordination with subnational government will be crucial for ensuring successful expenditure reform. Subnational governments often account for a sizeable share of the adjustment during successful fiscal consolidations (Darby, Muscatelli, and Roy, 2004; and Kumar, Leigh, and Plekhanov, 2007). Clarifying expenditure responsibilities and revenue assignments has helped strengthen budget constraints on local governments, while negotiation of binding fiscal targets has helped to coordinate policies across the different tiers of government. Use of cooperative arrangements between different levels of government also helps increase ownership of shared economic and fiscal objectives (IMF, 2009).

Age-Related Spending

Significant challenges lie ahead in dealing with age-related spending, especially health care. Public expenditure on pensions is projected to rise by 1 percentage point of GDP between 2010 and 2030 in the advanced economies. The relatively modest increase in this spending—in spite of the ageing of the population—reflects the significant reforms that have already been made in many countries. A further deepening of these reforms could place public pension spending on a sustainable path. In health, in contrast, the outlook is more challenging. Staff projects an increase in spending of about 3½ percentage points of GDP over the next 20 years. Containing the growth of public expenditures on health care will thus need to figure prominently in fiscal consolidation strategies over the next several years.

Pension systems

Staff projects that pension spending will increase by an average of 1 percentage point of GDP over the next 20 years (Appendix 3 and Figure 2.6). Large increases in pension expenditures are projected in advanced countries that have not substantially reformed their traditional pay-as-you-go systems (especially in Belgium, Greece, and Luxembourg). In other advanced economies, the increase in pension expenditures would be less marked due to the projected impact of already legislated reforms in offsetting the demographic pressures (Appendices 4 and 5).11 Adjustment needs may well be larger, though, as the projections assume that these reforms will not be reversed, even when they involve large cuts in replacement rates (as in Italy and Japan). Among the emerging economies, those with relatively high spending in 2010 are projected to experience the steepest increase in pension expenditures (especially Russia and Ukraine) over the next 20 years. In several other emerging countries, where coverage is currently low, the projected increase in expenditures is much less severe (China, India, and Pakistan).12 Beyond 2030, emerging economies are expected to experience a faster pace of ageing compared to the advanced economies.

Figure 2.6.Change in Public Pension Expenditures, 2010-30

(In percent of GDP)

Sources: Country authorities; EC (2009); OECD (2009); ILO (2010); and IMF staff estimates.

The cumulative fiscal cost of future pension spending increases is large (Table 2.2). Over the next 20 years, the net present value (NPV) of pension spending increases is about 8½ percent of GDP for advanced economies and 8 percent of GDP for emerging countries. The fiscal cost of pension increases over the subsequent 20 years is even larger—over 20 percent of GDP for both advanced countries and emerging economies.13

Table 2.2.Net Present Value of Future Pension Spending Increases(In percent of GDP)
2011-20302031-20502010-onwards
Average8.323.2153.5
Advanced8.721.5138.4
Emerging7.825.9177.1
G207.720.4129.6
Advanced7.316.7100.3
Emerging7.226.1174.3
Source: IMF staff estimates.
Source: IMF staff estimates.

Advanced and emerging economies face different challenges. In countries where coverage is extensive, the share of elderly population is larger, and spending is high—mainly the advanced countries—the primary objective should be to stabilize pension expenditures over the longer term while maintaining a reasonable rate of return on pension contributions and ensuring that pension benefits are adequate to prevent old-age poverty. In contrast, in the emerging economies, which generally have lower expenditures due to younger populations and less extensive coverage, the challenge is to expand pension coverage, but in a manner that does not generate fiscal imbalances as these systems mature.14 For emerging economies with high household savings rates (such as China), increased pension coverage would also support efforts to make domestic demand the primary catalyst of growth (see Baldacci and others, 2010).

Three policy options are available to offset the projected increase in spending of 1 percentage point of GDP between 2010 and 2030.15Figure 2.7 illustrates the tradeoffs across the typical options available to offset increases in pension spending—raising the statutory retirement age, reducing benefits, or increasing contribution rates. A two-year increase in the statutory retirement age would be sufficient to stabilize pension spending as a share of GDP at its 2010 level over the next two decades. This two-year increase in statutory retirement ages is roughly equivalent to a cut in benefits of 15 percent (corners of black solid line) and delivers similar fiscal effects as a 2 percentage-point increase in payroll taxes (origin).16

Figure 2.7.Illustration of Policy Options to Offset a Pension Spending Increase of 1 Percentage Point of GDP

Source: IMF staff estimates.

Raising statutory retirement ages should be the starting point for reform. Raising retirement ages would have a powerful effect: a one-year increase in the statutory age in the advanced countries would offset about half of the increase in spending projected between 2010 and 2030.17 Increases in statutory retirement ages are largely justified by the projected increase in longevity over the next 20 years: between 2010 and 2030, the number of years individuals are expected to live beyond the statutory retirement age is projected to increase by an average of 2 years in emerging and advanced countries (Table 2.3). Increases in the statutory retirement age should be accompanied by steps to limit the generosity of early retirement programs, which allow individuals to claim pensions, on average, by about 4 years earlier than the statutory age. It will also be important to tighten eligibility for disability pensions.

Table 2.3.Statutory Retirement Ages and Years in Retirement
Earliest eligibility age for pension benefits, 2010Statutory retirement age, 2010Life expectancy at statutory retirement age, 2010Life expectancy at statutory retirement age, 2030
Average58.963.017.919.9
Advanced60.164.217.719.7
Emerging57.161.218.220.3
G-2057.862.418.220.0
Advanced G-2060.464.018.520.3
Emerging G-2055.460.917.919.7
Source: IMF staff estimates.Notes: Legislated and planned increases in statutory retirement ages are included in the calculations for 2030. See Table A.4 for figures by country.
Source: IMF staff estimates.Notes: Legislated and planned increases in statutory retirement ages are included in the calculations for 2030. See Table A.4 for figures by country.

Raising the statutory retirement age, however, may not be sufficient in some countries to offset the projected increases in pension spending. The remainder of the increase in expenditures could be addressed with a combination of benefit reductions and increases in contributions.

  • Reduce benefits. Many advanced countries have already moved in this direction—in Japan, Korea, and Sweden, benefit cuts of nearly 20 percent or more are set to occur within the next 20 years (Table 2.4). Benefits could be reduced by modifying the base used to calculate benefits, modifying indexation rules, or taxing pensions.18 Cuts in pensions, however, should preserve benefits that are sufficient to lift the elderly out of poverty. Consideration should also be given to rules that link benefits and contributions to demographic and economic variables to maintain actuarial balance.19 Additionally, economies looking to expand coverage while containing the growth of expenditures might consider means testing of pensions (as in Australia, and to some degree Canada).20 Means testing, however, could weaken the link between contributions and benefits, hampering efforts to increase compliance and expand coverage.
  • Increase contributions. Changes in rates of social contributions need to be assessed together with potential changes in the rate of personal tax on labor income (discussed in Chapter 3), since it is their combination that determines the effective marginal and average tax rates likely to affect labor participation and hours worked decisions.21 Taxes on earnings are already high in a number of countries (in Austria, Belgium, France, Germany, Hungary, and Italy, the tax wedge is already near or above 50 percent of total labor costs). Other countries may have room for raising payroll contribution rates (Australia, Iceland, Ireland, Japan, Korea, New Zealand, and the United States have a tax wedge at or below 30 percent), and in some cases it may be appropriate to lift the ceiling on earnings subject to contributions. The incentive effects of social contributions, however, might be less marked if their payment is seen (correctly or not) as implying increased benefit entitlement.
Table 2.4.Tax Wedge and Replacement Rates
Social security contribution rate (in percent of labor cost)Total tax wedge (in percent of labor cost)Replacement rates
20102030Percent change, 2010-2030
Australia5.726.923.123.10
Austria36.548.854.254.20
Belgium34.256.046.547.93
Canada16.831.344.5
Czech Republic35.243.441.635.4-15
Denmark11.041.239.438.3-3
Finland24.343.551.251.71
France39.449.363.352.9-16
Germany33.452.050.445.9-9
Greece34.442.472.285.919
Hungary38.354.142.338.8-8
Iceland5.228.352.852.80
Ireland14.422.928.530.47
Italy31.546.571.364.1-10
Japan22.429.540.633.9-17
Korea15.820.357.846.2-20
Luxembourg22.535.941.439.3-5
Netherlands31.245.041.840.4-3
New Zealand0.021.241.141.10
Norway18.337.756.253.3-5
Poland33.739.759.645.1-24
Portugal28.137.649.042.3-14
Slovak Republic31.438.945.841.0-10
Spain28.037.862.661.0-3
Sweden29.844.648.136.6-24
Turkey29.339.786.969.5-20
United Kingdom18.032.834.634.50
United States14.330.138.735.0-10
Average24.438.549.545.9-7
Sources: OECD (2009a); and IMF staff estimates.
Sources: OECD (2009a); and IMF staff estimates.

Health care

Concerns about the sustainability of publicly-financed health systems have featured prominently in the United States, but much less in Europe; however, the outlook is grim also for Europe. Differences in assumptions about whether or not technological change will continue to drive up the cost of health care explain much of the differences in available projections for the United States and Europe. For the United States, the Congressional Budget Office (2007) projects an increase in health spending of 3.7 percentage points of GDP over the next two decades, based on the assumption that the increased spending per-capita arises from better, but also more expensive, medical services due to continued technological progress. In contrast, the European Commission’s Ageing Report (European Commission, 2009)—widely used for international comparisons—projects an increase in health spending of 0.7 percentage point of GDP, using a baseline assumption of no further increase in per capita spending due to technological progress. While much uncertainty exists, this is an extreme assumption that appears unrealistic based on historic trends.

Under the assumption that relative prices for health services will continue to rise in line with recent trends, staff projects that public spending on health will also continue to rise at a fast pace in both advanced and emerging economies. Public expenditures on health care are forecast to increase by over 3½ percentage points of GDP by 2030 in advanced countries, to 10½ percent of GDP on average (Figure 2.8).22 In emerging economies, the projected increase amounts to 1 percentage point of GDP, reaching 4 percent of GDP. In more than half of countries, public health spending would exceed 8 percent of GDP by 2030 (Table A.5). France, Germany, and the United States are projected to experience the largest increases of at least 3½ percentage points of GDP, while the smallest increases would be in India and Pakistan (less than ½ percentage point).

Figure 2.8.Increase in Public Health Spending in Baseline Scenario

(In percent of GDP)

Sources: IMF staff estimates; and sources listed in Appendix 6.

Ageing and other non-demographic factors will also contribute, albeit to a lesser extent than technology. These include income growth, the expansion of insurance, and provider reimbursement methods (Gerdtham and Jonsson, 2000; and Smith, Newhouse, and Freeland, 2009).23 In contrast to pensions, demographic change alone accounts for a relatively modest share of the projected increases in health spending (Figure 2.8). The relatively high increase in health spending—compared to pensions—also reflects the fact that the pension projections incorporate reforms that have already been agreed in legislation, which will help offset the effects of ageing. In health, in contrast, no major reforms have been agreed, or they are too uncertain to incorporate into the projections.24

The principal policy challenges differ in advanced and most emerging economies. In advanced economies, public health care systems are well developed, and the top priority is to contain the high rates of spending growth that have led to marked increases in spending-to-GDP ratios over the past 50 years (Table 2.5). In emerging economies, in contrast, the challenge is to expand basic coverage to a larger share of the population at a reasonable cost, without generating fiscal pressures. In these economies, the public system often provides coverage for a small share of the population, and in some cases, this coverage is insufficient to protect against the risk of illness among those covered.

Table 2.5.Public Health Expenditure in Advanced Economies(In percent of GDP)
196019701980199020002007Change, 1960-2007Change, 1970-2007
Australia1.93.13.94.65.56.04.12.9
Austria3.03.35.16.17.67.74.74.4
Belgium6.5
Canada2.34.85.36.66.27.14.82.3
Czech Republic4.65.95.8
Denmark6.67.96.96.88.21.6
Finland2.14.15.06.25.16.142.0
France2.44.15.66.48.08.76.34.6
Germany4.46.66.38.28.03.6
Greece2.33.33.54.75.83.5
Hungary6.34.95.2
Iceland2.03.15.56.87.77.75.74.6
Ireland2.84.16.84.44.66.13.32.0
Italy6.15.86.7
Japan1.83.24.74.66.26.64.83.4
Korea0.81.62.13.5
Luxembourg2.84.85.05.26.63.8
Mexico1.82.42.7
Netherlands4.15.15.45.07.31.4
New Zealand4.25.25.76.07.12.9
Norway2.24.05.96.36.97.55.33.5
Poland4.43.94.6
Portugal1.53.43.86.47.15.6
Slovak Republic4.95.2
Spain0.92.34.25.15.26.15.23.8
Sweden5.88.27.47.07.41.6
Switzerland4.35.66.4
Turkey0.71.63.14.1
United
Kingdom3.33.95.04.95.66.93.63.0
United States1.22.63.74.85.97.36.14.7
Average (PPP GDP weighted)5.53.9
Sources: OECD Health Database (2009d).Note: Data for actual or closest year available.
Sources: OECD Health Database (2009d).Note: Data for actual or closest year available.

Reforms of the health care systems will need to take into account the different mixes of public and private financing and service provision. For example, the United Kingdom and Italian systems comprise largely public financing and public provision; the Canadian and French systems are characterized by public financing and private provision; and the U.S. system is roughly split between public and private financing and mostly private provision. Across these different systems, both public and private health spending has increased as a percentage of GDP.

Various reforms to contain spending growth and/or improve the efficiency of spending could be considered. Past efforts in this area—including budget caps in a number of European countries in the 1980s, internal market reforms in the United Kingdom in the 1990s, and managed care in the United States in the 1990s—provide valuable lessons for future reforms (Appendix 7), although the appropriate policies will be country-specific, and depend on existing systems. Many of the reforms involve difficult tradeoffs, as they would result in a reduction in the quantity of services financed by the public sector. In light of the tremendous welfare gains produced by health advances (Murphy and Topel, 2006), the principal challenge will be to contain the growth of spending while ensuring broad access to high quality health care.

Supply-side

  • Reimburse providers using case-based payment or global budgets rather than fee-for-service. This option is important for both advanced and emerging economies. Fee-for-service, which is prevalent in both the United States and in Europe (Belgium, Denmark, France, Germany, and Luxembourg), gives physicians financial incentives to deliver additional services. Case-based payment methods, such as capitation and diagnosis-related groups that bundle different services into one lump sum, are an alternative. There is evidence that moving from fee-for-service to prospective payment can reduce expenditure between 10 and 20 percent (Eggleston and Yip, 2004). On average, switching from fee-for-service to prospective payment methods might reduce spending by 0.1 to 0.2 percent of GDP.25 To avoid adverse effects on health outcomes, mechanisms should be in place to ensure that providers do not reduce the quality of care or exclude less healthy patients. Greater use of the principles of supply-side control embodied in managed care (such as that provided by health maintenance organizations) is also an option for controlling costs while maintaining quality care (Cutler, McClellan, and Newhouse, 2000). Another option is to implement and maintain a hard budget constraint through a global cap on provider payments, which by construction will contain spending.26
  • Reduce the generosity of the publicly financed benefits package. This option is more relevant for advanced than emerging economies, as public health services are more generous in the former. These reforms would encourage the financing of some health care by the private sector, which already plays an important—but varying—role in all countries. For example, in Canada, most prescription drugs are not covered by public funds, but rather by private health insurance.
  • Strengthen evaluations of the cost-effectiveness of medical treatments and technology. In the short term, this is most relevant for advanced economies. Many countries (the United Kingdom, Australia, Netherlands, Sweden, and Finland) have established government bodies that assess the cost-effectiveness of new and existing technologies. Declining to pay for treatments that add small benefits at high incremental costs signals to the R&D sector to develop cost-effective technology. However, such a policy could also reduce the pace of innovation in some areas, which could lower dynamic efficiency (Jena and Philipson, 2007).
  • Implement health information technology (IT) to increase the efficiency of service delivery. The use of health IT varies widely across advanced economies. This could include, for example, improved data on patient histories (OECD, 2008b). In advanced countries, assuming public health spending is 6½ percent of GDP on average, widespread implementation of health IT could reduce spending by 0.2 percent of GDP.

Demand-side

  • Increase cost-sharing to discourage moral hazard. This option is more appropriate for advanced economies, where public health expenditure is at relatively high levels. Higher copayments or coinsurance rates for patients would shift some of the costs onto households and could help rationalize the utilization of health care services. A 5 percent increase in the share of cost patients absorb for outpatient care could, on average, reduce spending by 0.1 percent of GDP. The magnitude of any savings will depend on the extent to which other services are complements or substitutes.
  • Reduce tax expenditures for private health insurance. In countries where private health insurance contributions are exempt from taxation, favorable tax treatment should be reconsidered. The size of these tax expenditures can be large, and some argue this subsidy leads to “overinsurance” (Feldstein, 1973). This issue is most often discussed in the U.S. context—where these benefits amount to about 2 percent of GDP—in light of its employment-based, private insurance system. However, subsidies for private insurance also exist in Australia, Denmark, and Greece.

    Other Resources Citing This Publication