Sweden
Selected Issues
Author: International Monetary Fund
Publication Date:
09 Aug 2004
eISBN:
9781451835960
Language:
English
Keywords:
ISCR; CR; sickness absence; absence rate; insurance system; replacement rate; cash benefit; personal income; government structure; tax revenue; tax rate increase; Sweden's government sector; Public employment; Europe; government expenditure
Download PDF (1.9 MB)

The Selected Issues paper analyzes tax policy trends at the local level in Sweden and assesses the effectiveness of the vertical fiscal policy coordination system. The study reviews Sweden’s local public finances from an international perspective and empirically explores various explanations for the gradual increase in local tax rates. It discusses long-term challenges for local public finances and aggregate fiscal policy coordination. The design of vertical fiscal policy coordination in other countries is described. The paper also examines the Swedish experience of work absence in a European context.

Abstract

The Selected Issues paper analyzes tax policy trends at the local level in Sweden and assesses the effectiveness of the vertical fiscal policy coordination system. The study reviews Sweden’s local public finances from an international perspective and empirically explores various explanations for the gradual increase in local tax rates. It discusses long-term challenges for local public finances and aggregate fiscal policy coordination. The design of vertical fiscal policy coordination in other countries is described. The paper also examines the Swedish experience of work absence in a European context.

I. Fiscal Policy in a Decentralized Economy1

A. Introduction

1. Sweden’s government sector is exceptionally large and highly decentralized. Among the EU-15 member countries Sweden’s general government expenditure ranks the highest in terms of GDP. At the same time the delivery systems for public services are highly decentralized. More than 40 percent of general government spending is carried out by local governments—20 county councils and 290 municipal governments—pertaining primarily to health care, long-term and disability care, education, child-care, and other services. Own revenue sources at the subnational level cover about 80 percent of expenditures.

2. The broad distribution of responsibilities across government levels poses a formidable challenge for macro-fiscal management. As local governments have autonomy over a wide range of issues including tax policy, fiscal policy needs to be well coordinated to ensure economic stability and fiscal sustainability. This requires an effective policy coordination mechanism that brings budgets of local governments in line with aggregate policy objectives. The two most common coordination mechanisms among European countries are a cooperation approach where all participants are involved in policy formulation and execution, or the use of fiscal rules that prescribe policy targets for the different levels of governments. Sweden has opted for the latter approach.

3. Sweden’s current fiscal policy framework addresses the need for vertical coordination through a three part fiscal framework. For the general government, Sweden targets a structural surplus of two percent of GDP. The surplus target defines the overall policy direction and anchors a rolling medium-term budget. The second component is an expenditure rule for the central government. Aggregate expenditure is constrained by nominal expenditure ceilings set for three years. The ceilings include budget margins to allow for cyclical variations in spending and to give some flexibility in the budget planning process. The third element is a balanced budget requirement for local government. Any deviations from this target need to be offset within two years.

4. The fiscal framework has so far been very effective. With a general government surplus of 0.5 percent of GDP in 2003 Sweden’s fiscal position remains favorable in a comparative EU context. The government also managed to reduce the overall tax burden by 1.9 percentage points of GDP since the inception of the new fiscal framework. That said, fiscal policy has benefited from relatively advantageous economic conditions in the late 1990s, which provided the fiscal room to react aggressively to the economic woes from restructuring of the telecommunications sector and the bursting of the technology bubble. While the economy weathered the recent global downturn relatively well, efforts have visible slackened to return to the general government surplus target of 2 percent by 2006, as initially envisaged.

5. One sign of the strains in the fiscal framework are persistent increases in local tax rates. Over the last ten years local income tax rates have steadily increased.2 Since local tax revenue is cyclically sensitive and transfers from the central government do not automatically adjust to the business cycle, local governments have struggled to meet the balance budget requirement. As a result, tax rates have begun to creep up, in particular since the end of the 1990s. Moreover, local spending appears to have followed a procyclical pattern, which has put a disproportionate burden for countercyclical fiscal policy on the central government

6. The objective of this paper is to analyze tax policy trends at the local level and to assess the effectiveness of the vertical fiscal policy coordination system. The study reviews Sweden’s local public finances from an international perspective and empirically explores various explanations for the gradual increase in local tax rates. It further discusses long-term challenges for local public finances and aggregate fiscal policy coordination. It describes the design of vertical fiscal policy coordination in other countries and underscores country experiences relevant to Sweden.

7. The paper is structured as follows. The second section presents an overview of the institutional government structure in Sweden. The discussion highlights characteristics of Swedish local public finances in an international perspective, and points out the main differences to European comparator countries. The third section focuses on the recent trend increase in local tax rates. It examines various possible explanations and provides econometric evidence on likely determinants. It concludes by discussing longer-term expenditure challenges for local governments. The next section gives an overview of alternative vertical fiscal coordination mechanisms and singles out individual country experiences relevant for Sweden. The final section concludes.

B. Sweden’s Local Government in International Comparison

8. Sweden has a long-standing tradition of local and regional self-government. The principle of self-government is enshrined in the Swedish constitution and the state has assigned broad responsibilities to local authorities for the provision of public services. Rights and obligations for local governments are laid out in the Local Government Act (LGA), which was enacted in 1992.

The main subnational government bodies are:

  • a total of 290 municipalities, which are mainly responsible for the provision of primary and secondary school education, childcare, long-term care of the elderly, disability care, and other social and local services. The average population size of municipalities in 2003 was around 31.000 with the median size at around 15.000 inhabitants. The majority of smaller municipalities are in the Northern part and tend to have a below average per capita tax base.

  • 20 county councils in charge of health and medical services.

9. Municipal and county councils are elected bodies and are required to provide public services on a balanced budget basis. Local governments are mandated to carry out public services assigned to them by law,3 but have the right to raise the necessary revenue to finance their activities. Municipalities have no restrictions on borrowing, and the central government plays no part in either monitoring or approving local government accounts. Since 2000, municipal and county council budgets are required not to show a deficit ex ante (LGA 8/4). If ex post expenditures for a particular financial year exceed income, then the deficit must be offset in the budgets of the subsequent two years. In exceptional cases, which are not defined by the law, local governments can decide that no such adjustment be made (LGA 8/6). The LGA does not contain sanctions for failing to meet the balanced budget or deficit adjustment requirement. Changes in the balance budget requirement has been proposed in a bill to the parliament in May 2004. The changes are expected to come into effect by January 2005. Municipal and county council budgets may then, if special circumstances are at hand, show a deficit ex ante. Ex post deficits are to be offset in the budgets of the subsequent three years.

Size and composition of subnational spending

10. Subnational spending in Sweden exceeds the average expenditure level of European comparator countries. Figure I.1 compares the share of local government expenditure in general government expenditure for various European countries in 1985 and 2001. In Sweden subnational government spending reached 43 percent in 2001 and was only exceeded by Denmark and Canada. This high level is in line with other Nordic countries which traditionally rely on local governments to provide key public services. More generally, in most European countries devolution of public services has gradually moved forward during the 1990s. This trend has however been uneven. Decentralization of public services in unitary states became visibly more pronounced, and caught up to levels in federal states.

Figure I.1.
Figure I.1.

Sweden Subnational Spending as a Share of General Government Spending

Citation: IMF Staff Country Reports 2004, 245; 10.5089/9781451835960.002.A001

Source: OECD 2003

11. Differing growth trends in subnational spending between unitary and federal states are of particular interest due to their implications for vertical policy coordination. Federal countries have traditionally placed more emphasis on a cooperative approach on fiscal policy partly because their government structure by definition required more intensive intergovernmental collaboration. In Germany for instance, medium-term fiscal policy targets are based on voluntary agreements between the Länder governments and the central government. An attempt to introduce a more formal arrangement failed in the run-up the EMU (Wendorff 2001). In unitary countries, on the other hand, a more formal approach has been adopted (e.g., Spain adopted a fiscal responsibility law in 2003). As unitary states have expanded the scope of subnational public finances, issues of fiscal relations among government levels have gained more prominence. These challenges are explored in more detail in section E below.

12. The composition of Sweden’s local expenditure is tilted towards demographically sensitive areas. Similar to most European countries, Sweden’s subnational responsibilities are broadly spread and cover most aspects of public services. Sweden stands out in a comparison with other unitary states with its high emphasis on health and social services. Taken together these two items account for over 50 percent of subnational spending, while in comparable unitary states they cover less than 30 percent. Not surprisingly, expenditure pressures from aging related areas have been of great concern to local governments in Sweden.

Sweden: Composition of Subnational Spending

article image
Source OECD (2003)

unweighted average

13. An added problem with the devolution of health care related services is an unclear separation of tasks between municipalities and county councils. In the early 1990s the government transferred responsibilities for long-term and disability care from county councils to municipalities. While the expenditure assignment was motivated by a concern to improve care through closer local accountability, the separation has led to implementation problems for two reasons. First, long-term care and disability services can often not be cleanly separated from health care related tasks, which has led to confusions about responsibilities. Second, the separation of functions has also imposed a formidable coordination problem, requiring a frequent information exchange between a large number of municipalities and the relatively small group of county councils. Local administrations have recognized these weaknesses. To improve coordination the two representation bodies for municipalities and country councils are being merged into one unified agency dealing with both municipal and county issues.

Revenue assignments

14. The income tax is the largest revenue item and the only tax revenue source of local governments. In 2003 the local income tax amounted to 16.7 percent of GDP compared to 27.3 percent of tax revenue by the central government. By law, income from employment is subject to both local and national taxation. Local taxes are proportional and levied separately by municipalities and counties. In 2003 local tax rates varied across municipalities between 28.9 and 33.7 per cent, with an average (population weighted) rate of 31.2 per cent. Taxable income includes salaries and pensions as well as fringe benefits although deductions are possible for certain work-related expenses. A relatively low allowance is provided which varies between SEK 11 400 and SEK 25 900 depending on income. Tax credits are granted for 25 per cent of union dues and 40 per cent of unemployment insurance fees. National income taxes are levied at 20 and 25 percent with high thresholds so that less than 10 percent of all full-time employees pay any national income tax.4

15. Other own revenue sources are small. Local authorities are permitted to charge user fees for services they provide. They can also engage in entrepreneurial activities and many municipalities own public enterprises. Income from fees for services (e.g., child care, elderly care, and health and medical care) and other non-tax income (e.g., interest income) account for 10 percent of subnational revenue

16. Government transfers to local governments are divided in roughly equal parts into block and specific grants. In 1993, a reform was initiated to permit greater flexibility and autonomy for local governments. As a first step, a large number of specific grants were replaced by a single new block grant, which was distributed according to local tax capacity and structural costs. A new per capita based central government grant was added in 1996 to further decrease reliance on specific grants. By 2003, vertical block grants amounted to roughly one half of overall transfers to local authorities. The remaining specific grants are used mainly for financing spending on pharmaceuticals, elementary and secondary education, and childcare services. In 1993, the “financing principle” was also introduced. Any measures introduced by the central government that directly affects local authorities must be accompanied by a means of financing that does not involve raising local taxes.

17. The level of vertical grants are not determined by a formula, but decided by the central government during the annual budget process. The central government has full authority to decide over the level of transfers to local governments within the requirement of the “financing principle.” While formal consultations are held with local governments, there is no fixed rule or a mechanism that links changes in compensation to trends in local government demands or economic developments.5 This uncertainty about the level of future vertical grants has hampered the ability of local governments to devise realistic medium-term plans and has even affected finances for the current budget year. One recent example is the larger than expected reduction in vertical grants in compensation for a tax base broadening measure related to the taxation of pensions.

18. The lack of local tax base diversification is partly an unavoidable result of Sweden’s large local government sector. Public finance theory suggests that stable and immobile tax bases should be assigned to local governments (e.g., Mueller 2004) to prevent excessive volatility and tax competition among local governments. Most commonly, this entails the assignment of property taxes or excises to local governments. In Sweden, revenue from both of these taxes accrue to the central government. However given their small revenue capacity relative to financing needs—property taxes and excises amount to less than 5 percent of GDP—they would not be an substitute to the personal income tax. In fact, all Nordic countries rely heavily on the income tax as a the main local revenue source (text figure). Countries with a more balanced tax base distribution, such as France, Spain or Portugal, have either significantly smaller local governments or rely to a larger extent on vertical government grants.

uA01fig01

Composition of Local Tax Revenue: 2001

(in percent)

Citation: IMF Staff Country Reports 2004, 245; 10.5089/9781451835960.002.A001

19. Sweden’s local governments have very high tax policy autonomy. The right of local authorities to levy taxes is set out in the constitution. In practice this means that local authorities can decide on the local tax rates while the state defines which tax base should reside with the local authorities. In an international comparison this high degree of local tax policy autonomy stands out. The text figure depicts two measures of discretional tax setting power. The two vertical bars measure the share of local taxes—of total tax revenue or of GDP—for which local authorities have either full control over the tax base or the tax rate. Sweden scores very high on both indicators. Even among Nordic countries this high degree of autonomy is unusual.

uA01fig02

Local Tax Policy Autonomy

(share of local taxes with full local control over tax rate or base)

Citation: IMF Staff Country Reports 2004, 245; 10.5089/9781451835960.002.A001

C. Local Tax Rate Structure and Trends

20. Statutory tax rates of local governments have steadily increased since the early 1990s. The average (population weighted) personal income tax rate in 2004 is 31.5 percent, roughly 1.6 percentage points higher than in 1993.6 The text figure below plots the average and the population weighted level of local tax rates. Both measures show the same upward trend with a slight narrowing between the two at the end of the observation period. The text graph also indicates that the tax hike occurred in two main stages, one in 1994–96 and another one beginning in 1999. These two shifts can be partly explained by institutional changes which are discussed below.

21. The local tax rate increase has been of macroeconomic significance. In 2003 local tax revenue reached 16.7 percent of GDP, up 2.2 percentage points compared to the lowest revenue ratio recorded in 1995. The upward tax creep since 1999 eliminated more than one quarter of the tax relief granted by the central government in 2000 and 2001. A simple rule of thumb for the revenue impact of tax hikes can be derived by dividing the local tax revenue by the statutory tax rates. Assuming a revenue elasticity of 1, a one percentage point increase in the local tax rate raises revenue by 0.8-1.1 percentage points of GDP. Since local governments expect further tax rate increases in the coming years—between 1-2 percentage points by 2006—the tax burden would increase significantly.

uA01fig03

Sweden. Average and population weighted local income tax rates: 1993-2004

Citation: IMF Staff Country Reports 2004, 245; 10.5089/9781451835960.002.A001

22. The increase in tax rates has been broad based, despite various attempts to regulate or influence local tax policy. Between 1993 and 2003 local tax rates increased in all but 7 municipalities. The box plot shows the distribution of local tax rates over time. As the graph clearly demonstrates, tax rates have gone up in all parts of the tax rate spectrum. Moreover since 1999 a narrowing in the tax differential between low tax and high tax localities has occurred. The general upward drift in tax rates is quite remarkable in its uniformity given different attempts to contain local tax rates. During 1991-1993 a constitutional regulation prohibited a rise in the local income tax. After this regulation expired in 1994, the government offered local governments a compensation if they maintained the level of taxation. Despite these incentives the prolonged tax freeze led to a sharp increases in tax rates in 1995 and 1996. A new attempt in controlling local tax rates was made in 1997-99. During this time local governments received lower grants if their tax rate increased. This policy led to an aggregate decline in tax rates, but was abandoned. By 2000 tax rates began to rise again with a significant acceleration beginning in 2002. The latter increase may be related to the introduction of the balanced budget requirement in 2000.

uA01fig04

Sweden. Box plot of local tax rates 1993-2004

Citation: IMF Staff Country Reports 2004, 245; 10.5089/9781451835960.002.A001

Shaded box identifies range between 25th and 75th percentile.Horizontal line within box indicates median tax rate.excludes outside values

23. The dispersion of local tax rates has been quite low reflecting effective revenue equalization. The difference between the lowest and the highest local tax rate is 5.2 percentage points in 2004. Moreover the tax rate structure across municipalities was stable over this period. The majority of local governments which had tax rates in the lowest quartile of tax rates distribution in 1980 was still in this group by 2000, and vice versa for local governments with high tax rates. Part of the reason is an extensive revenue equalization system. The scheme reallocates funds from regions with high tax bases to low tax base areas, thus reducing the need for compensatory tax rate differentials. In addition, local government receive compensation for cost differences further lowering the need for a greater tax rate dispersion. In 2003 income and cost equalization schemes redistributed about 1.1 percent of GDP of local revenue between local governments.

Determinants of local tax rate structure

24. Local tax rates are higher and tax bases lower in smaller municipalities. Table I.1) describes the distribution of tax rates and disposable per capita income across municipalities sorted by population size. Local tax rates in smaller municipalities with a population of less than 50,000 are between 1 and 1.7 percentage points higher than in large cities (> 250.000). Part of the reason is that larger communities have higher tax bases. The average per capita tax base in the largest municipalities is about 20 percent higher than in the smallest ones. A large part of this discrepancy is however offset through the revenue equalization scheme and balancing grants. Thus other factors such as the population age structure and the expense ratios for other large local public services must be important.

Table I.1.

Sweden: Local Tax Rates and Tax Base by Municipalities (2002)

article image
Source: Statistics Sweden

differences between group averages statistically significant at the 1 percent level.

Table I.2.

Sweden: Determinants of the Local Tax Rate Structure 1/

article image
Absolute value of τ statistics in parentheses; * significant at 5%; ** significant at 1%

dependent variable: level of statutory local tax rate

assessed personal income municipal population

annual expense per municipal population

25. A more systematic analysis reveals that municipal income levels and per capita expenses for public services determine the tax rate structure across local governments. A pooled cross sectional analysis was carried out on municipal level data for the last two available years 2001 and 2002. The dependent variable is the level of statutory local tax rates τ (county plus municipal tax) in a municipality. The estimation method is ordinary least squares and controls for municipal population size, age structure, and levels of per capita expenses on three types of public services (education, childcare, and long-term and disability care). Municipal level data on expense levels for health care were not available and were approximated by the old-age ratio of local population.

26. The following findings can be derived from the static analysis of the tax rate structure. As on would expect, the local tax base size is negative related to tax rates in all regressions. A one standard deviation higher tax base is related to a 0.3–0.5 percentage points lower tax rate assuming that all other variables remain unchanged. An older population structure and higher per capita expenses for education and long-term and disability care are unequivocally associated with higher local tax rates. The effect of higher expenses on childcare is negative (model 1), but turns out to be affected by an omitted variable bias. Once the level of long-term care expenditure are taken into consideration (model 5), the negative sign of the childcare cost vanishes indicating a negative correlation between the level of local childcare and elderly care costs The level of health care expenditures are approximated by the share of elderly population in a municipality and is estimated to also increase local tax rates (model 3). In model 5 its effects turns negative due to multicollinearity with the elderly care cost variable. Among the different expense measures, the largest impact stems from long-term care expenses. A one-standard deviation higher expenditure level translates into a 0.75 percentage points higher local tax rate. It is also interesting to note that once all variables are included in the model, local population size does not matter. In other words, there is no big city advantage per se.

Determinants of local tax rate increases

27. The recent increase in local tax rates can be attributed to a number of different factors. Potential reasons for the tax rate increase are: an erosion of the income tax base, cyclically induced ratcheting up of public consumption, and emergence of underfunded expenditure mandates:

  • Tax base erosion. A decline of the tax-base due to lower employment rates or moderate wage growth could have contributed to financing shortfalls and forced local government’s to raise revenue through higher tax rates.

  • Ratcheting up of public consumption. Local governments may have raised local public consumption during favorable economic situations and struggled to reduce spending during downturns, eventually requiring them to raise taxes.

  • Emergence of underfunded mandates. Demographic changes as well as a gradual expansion of the welfare benefit system may have led to a disproportional increase in the demand for public services. To the extent that local expenditure needs did not grow in line with available financing sources, local governments may have been forced to raise tax rates.

Competition over the local tax base appears to not have halted the tide towards raising tax rates. Interregional labor mobility is low in Sweden and migration decisions appear unrelated to the level of local tax rates (Aronsson, Lundberg and Wikstrom 2001). The plausibility of the three explanations are discussed below.

28. There is little evidence of local tax base erosion. Various local tax base measures show that disposable income has grown in line or even exceeded GDP growth. The text figure shows a box plot of the distribution of real percapita taxable income. Average municipal income has grown steadily since 1993 and even increased in percent of GDP in the last three years. Similarly, aggregate local tax revenue divided by the statutory tax rate shows an upward trend if expressed in percent of GDP (not shown). While tax base broadening measures may explain this effect, the overall consensus is that the overall growth of the income tax base has been quite strong in the last decade. This effect is also confirmed in the multivariate analysis below.

29. A ratcheting up of public consumption may have contributed to rising tax rates, but its effect is difficult to identify. Some intuitive support for this argument can be derived from the text figure. At an aggregate level, real public consumption by local governments increased in two episodes, between 1993-1995 and 1997-2002 (dotted line). During both periods the acceleration of public consumption growth coincided with an improvement of the output gap (solid line), while the intermittent deceleration of consumption growth (1995-1997 and 2002-2003) coincided with a widening of the output gap. This sequence would be consistent with a cyclically induced ratcheting up of consumption. The associated rising financing needs then triggered local tax rate increases starting in 1999. In other words, poor expenditure discipline in good time contributed to the tax hikes. A recent study by Soderstrom (2003) indirectly supports this line of argument. While the aggregate trends intuitively fit this explanation, a more definite answer have to await a more comprehensive study of local fiscal behavior.

30. The effects of expenditure demand pressures were assessed through regression analysis. The model relates changes in local tax rates between 19957 and 2002 to changes in several basic characteristics (B) and expense indicators (E) for various public services:

Δτi=f(τ94i,ΔBi,ΔEi)    i…municipality

where Δ refers to the difference operator (∆X=X2002-X1995), and τi to the statutory tax rate in municipality i. By looking at tax rate changes over a longer horizon, the estimation should not be affected by temporary factors, such as the incentive scheme for controlling tax rate growth (1997-1999). The model is specified in time differences (∆), but includes the level of the local tax rate at the beginning of the period to control for differences in initial conditions. The three basic characteristics B are: population size, per capita amount of vertical grants, and per capita level of the local tax base. The expense variables E are municipalities’ per capita outlays for childcare, education, and long-term and disability care for the elderly. Data on health expenditures were not available and are approximated by the share of old age population in a municipality.

31.Results from multivariate analysis show that rising costs related to population aging have been a main factor for tax rate increases. The main findings from the regression analysis are:

  • There is no evidence that the initial level of tax rate in 1994 affects the tax rate dynamics. The negative sign in models 5 and 6 indicate some convergence to the mean, but its effect is statistically insignificant.

  • Rising expenses for long-term and disability care and an aging municipal population have significantly increased local tax rates. A one standard deviation higher expense ratio for long-term care (1866 SEK per capita) increased the tax rate by 0.2 percentage points. A one standard deviation increase in the population-age share has a more moderate effect of 0.05 percentage points.

  • Child-care or education expenses have not affected tax rate increases in a statistically significant sense.

  • Local population growth has helped contain tax rate increases. A population increase by 5,000 residents lowered the tax rate by 0.2 percentage points. The effect is probably driven by the net revenue effect from local immigrants who are in the active labor force.

  • As expected, growth of per capita tax base reduced the tax rate. The effect is however not statistically significant. This may indicate that an increase in the tax base may have induced local governments to increase public consumption and thus did not slow down the tax rate increase.

  • Higher grants from the central government have helped contain the tax rate growth.

  • An increase of the grants by 1000 SEK per capita decreases the tax rate by 0.7 percentage points leaving all other variables unchanged.

  • The findings are robust to the inclusion of a large-city dummy or controlling for the initial level of per-capita income, grants, or population.

Table I.3.

Sweden: Determinants of Local Tax Rate Changes Between 2002 and 1995 1/

article image
Source: staff estimatesAbsolute value of τ statistics in parentheses;* significant at 5%; ** significant at 1%

dependent variable: change in statutory tax rate between 1995 and 2002; Changes over time are expressed by the difference operator: ∆X = X02- X95;

2/ Taxable personal income per municipal population3/ Annual expense per municipal population

32. In sum, the regression results indicate that expenditure pressures have been a major factor for local tax rate increases. In particular, expenses on long-term care and disability care, and to a lesser extent an aging local population have been the main culprits. Anecdotal evidence suggests that part of the expenditure pressure is due to higher utilization of existing programs in particular in the case of disability benefits. It would therefore be too early to infer from the above findings that tax rate increases are due to demographic shifts. Another caveat is that the regression could not draw on detailed information on health expenditure trends, which exhibited large nominal cost increases.

D. Longer-Term Challenges for the Local Public Finances

33. Sweden’s local public sector will be strained by demographic changes. Population aging is affecting Sweden somewhat earlier than other EU countries. The old-age dependency ratio is currently at 30 percent—slightly above the EU 15 average—and will gradually increase to 44 percent in 2040 after which will stabilize at a lower level.8 While Sweden has early on addressed the financing needs of its public pension scheme, it has been less forward looking in ensuring the financial viability of health and long-term care programs. Population aging will lead to rising demands for these services and growing pressure on local public finances. However, population aging also means a smaller cohort of children entering childcare and subsequently education, which could offset some of these cost increases. Finally aging also leads to a decline in the share of the domestically born labor force with the implication that a steady labor supply would have to come from greater employment of foreign born residents (OECD 2004).

34.The text figure above depicts projected increases in the demand for public services stemming from compositional changes in population age structure. The projections distinguish between different public services provided by local governments and assumes constant utilization rates by age group and gender. Comparisons are made relative to the level of local public services provided in 2000 (standardized at 100). The depicted changes in the demand for public services over time are derived solely from the effect of different age and gender cohort sizes.

35. Over the next 20 years two main developments can be distinguished.

  • As fewer children are born, demand for childcare and education services will decline. The use of childcare bottoms out by 2005, but subsequently immigration gradually increases demand over the next 15 years. Demand for schooling exhibits the same time-delayed pattern, and will decline until 2015 when a turning point is reached. By 2020 about the same volume of primary education will be needed as in 2000.

  • The flip side of lower expenses for the young are rising demands for health related services. The need for both health and long-term care services will continuously increase, as more people grow older. By 2020 the volume of health and long-term care is expected to increase by 15 and 20 percent relative to 2000 levels solely due to population aging. The projections are based on an “unchanged-policy” basis and thus do not incorporate savings from potential reform measures. They do however also not account for higher utilization rates of existing services unrelated to population aging, such as greater familiarization with benefit entitlements under the existing system—a phenomenon currently observed for disability care benefits.

36. The main increase of expenditures will be driven by cost developments. Official estimates of the annual increases in service costs at constant volumes range between 4-5 percent per year. The text table summarizes the average expected cost developments for the respective local public services. Under the current generous benefit system the estimates for long-term care and health services appear conservative. Thus a tremendous financing needs could arise to local governments from both cost and demand increases.

article image

37. Financing problems from rising expenditure needs will be compounded by a shrinking labor force. Population aging decreases the labor force and thus reduces the available tax base for local governments. As Sweden has already a high labor force participation rate, in particular in pre-retirement cohorts, there is only little room to increase employment of the existing labor force. One possible alternative is immigration. Long-term demographic projections show that the share of foreign born population may increase from 11.8 percent in 2002 to about 16.5 percent by 2020. However immigrants have a lower employment rate, and OECD projections show that without further efforts to better integrate immigrants into the labor market, total hours worked may stagnate or only grow modestly until 2020.

38. Based on this outlook, there is little prospect for a tapering off of local tax rate increases. As local governments will face a growing gap between resources from weakening tax bases and rising expenditure needs, they will be forced to continue raising tax rates unless they receive greater financial support from the central government, introduce service fees, or reduce benefits. A particular concern of a rising aggregate tax burden would be its negative effects on employment and thus on the prospects for integrating the foreign born population into the labor market. While immediate expenditure pressure may be subdued due to declining childcare and education needs, local governments may have to raise tax rates in the near future to avoid larger increases after 2010, when the costs of population aging begin to bite. The Swedish Local Government Association estimates that an upfront savings strategy may require a 2 percentage points tax rate increase by 2006.

E. An International Perspective on Vertical Fiscal Policy Coordination

39. Fiscal decentralization poses a formidable challenge for macroeconomic management. Its overriding objectives are to maintain macroeconomic stability and to ensure fiscal sustainability. Both goals require that governments secure aggregate fiscal discipline and implement policies consistent with economic and demographic prospects. In decentralized economies this means that all tiers of government need to coordinate their fiscal plans and synchronize actions. Experience from a number of countries shows however that governments in decentralized economies have often not succeeded in achieving these goals due to ineffective fiscal management structures.

40. In decentralized economies, threats to fiscal discipline and sustainability arise from fiscal externalities, unclear assignments of responsibilities, and underfunded mandates. Subnational governments often do not incorporate the spillover effects of their decision making into their assessments. Costs of excessive spending and borrowing go beyond the local realm and are borne by higher borrowing costs for the rest of the economy. In the same vein, bail-out guarantees can induce moral hazard and may encourage fiscal laxity with similar consequences. A different problem is the emergence of underfunded mandates caused by transferring spending responsibilities to local authorities without fully taking costs into account. Examples of this burden shifting have been decentralization processes in Denmark, Norway, Russia, or the United States. Often emergence of underfinanced mandates reflect a poor information exchange between the center and the regions, but can also be the result of devolving fiscal problems to local authorities. Whatever the source of the underlying problem, the existence of coordination and enforcement mechanisms is critical to ensure fiscal discipline.

41. In Europe, efforts to improve vertical fiscal policy coordination have received increased attention with the introduction of the EMU. To ensure that fiscal policy supports the common currency, EU member countries subscribed to the Stability and Growth Pact (SGP). Under this arrangement countries commit themselves to avoid excessive general government deficits over 3 percent of GDP and to pursue the medium-term objective of budgetary positions close to balance. For many countries with a centralized government structure this merely meant adherence to new fiscal policy targets. However, for countries with a decentralized government structure the new restrictions imposed indirectly commitments on local authorities outside central government control.

42. In order to meet the new fiscal policy limits, many EU governments called for the introduction of internal stability pacts (see Banca D’Italia 2001). These internal pacts (ISP) are the national policy instruments that translate national fiscal policy goals into subtargets for the different tiers of governments. Formulation and success in implementing ISPs has been quite varied (see Daban et al 2001). In Spain, the ISP was initially based on negotiated fiscal targets for regions and the central government with varied success in implementation. In 2003 new legislation was passed setting tighter standards for greater transparency and implementation and has so far been successful in meeting fiscal targets. In Italy, efforts to impose fiscal discipline on regional governments through an ISP had only mixed success and faltered in recent years. In Austria the government relied quite successfully on recurring binding contracts between the federal and local governments, while Germany based fiscal policy on voluntary agreements between the central government and the Länder governments (Box I.1)

43. Two broad approaches for fiscal vertical coordination have emerged in the advanced economies. These are rules-based approaches relying on setting targets or limits, and cooperative approaches based on mutually agreed policies. The text table gives an overview of the most common practices used in achieving aggregate fiscal discipline (OECD 2003). Fiscal rules, such as the requirement of a balanced budget, or tight administrative controls have been adopted primarily in countries with a unitary government structure, such as Sweden. Cooperative approaches are more common in federal countries. They can be voluntary and indicative or entail a detailed fiscal contract as the result of a negotiation process. The distinction between the two strategies is gradual with significant differences in terms of tightness of rules and the formality of the cooperative process.

article image

44. The choice between the two coordination strategies reflects partly the degree of vertical power sharing. To the extent that subnational governments enjoy larger economic bases and political power, countries have relied on the cooperative approach to impose fiscal discipline. On the other hand, countries with a uniform government structure and tighter central control have relied more often than not on a rules-based approach. Two contrasting examples are Germany and France, with the former having powerful subnational governments (Länder), which made the use of a rules-based approach infeasible. In France a centralized government structure goes hand in hand with tight administrative controls over local governments. But even with a strong centre based government structure, the scope for rules based approach may be limited.

Cooperative Fiscal Coordination in Germany and Austria

Germany and Austria are both federal countries and use a cooperative approach to vertical fiscal policy coordination. While Germany has relied on setting non-binding policy targets, Austria uses a contractual approach, which defines jointly vertical and horizontal transfers, local and central fiscal targets policy, and financial sanctions for non-observance.

Germany’s vertical fiscal coordination process is conducted through an intergovernmental body (Finanzplanungsrat) chaired by the finance minister. The committee meets twice a year and is responsible for preparing fiscal projections for the general government. Targets are set within a medium-term horizon and based on independently derived tax revenue forecasts (Arbeitskreis Steuerschaetzung). In practice the fiscal coordination committee functions as a forum for reconciling competing fiscal plans between the regional and local governments (Gemeinden and Länder) on the one hand and the Ministry of Finance on the other hand. The determination of vertical and or equalization grants is made independently from this process and is based on revenue sharing rules. The committee has no formal authority and its targets are for the most part indicative.1 As Germany entered EMU, the fiscal coordination committee has become the substitute for a formal internal stability pact (Wendorff 2002) and its targets are consistent with Germany’s Stability Program under the SGP.

In Austria, vertical fiscal coordination is implemented through a fiscal equalization contract signed every four years in the context of negotiations about vertical and horizontal financial transfers. The agreement sets fiscal balance targets for all levels of government and determines financial sanctions. According to the last pact agreed upon in 2001, the regions have to produce an aggregate surplus of 0.75 percent of GDP per year with specific targets set for each individual region. Municipalities have to balance their budget. The targets are binding for each individual region as well as for the municipalities as a group within each region. Trading of deficit/surplus rights and obligations between local governments is possible. If governments fail to meet these targets, fines can be applied, but require a unanimous decision of a multi-government commission. Sanctions do not apply under serious economic conditions. The quantitative goals within the pact are annually reviewed and revised if necessary.

1 In 2003 expenditure projections were conceived as binding spending caps, but have not been adhered to.

45. The success of cooperative versus rules based approaches also depends on the political cohesion of the government. As Hallerberg, Strauch and von Hagen (2001) have convincingly argued, the type of political system plays an important role for the effectiveness of budgetary institutions and thus fiscal discipline. The study examines fiscal consolidations in Western Europe and argues that successful episodes of reform appear linked to the right mix of political and institutional conditions. In particular, different political constellations—one-party majority governments, multi-party majority coalitions, and minority governments—require different policy coordination mechanisms to be effective. Majority governments due to their strong support can rely on delegation schemes, such as a strong Ministry of Finance, to enforce policies. On the other hand, multi-party coalitions and minority governments are more likely to be effective under a commitment based scheme which binds together differing views of coalition partners and the opposition in detailed agreements (e.g., the Netherlands under the Kok administration). A corollary of this argument is, that countries with a broad political consensus will better be attuned to adhere to a fiscal rule, while countries with a less homogenous view and power bases in different regions will have to rely on carefully crafted cooperation agreements to ensure fiscal discipline.

46. The case of Denmark provides valuable experience for Sweden since it is both characterized by a large local government sector and minority governments. Similar to Sweden, the Danish government has devolved a wide variety of responsibilities to local governments. Local public sector services include childcare education, health and long-term care services and account for about 70 percent of total public consumption. In addition governments in Denmark had to rely on the support of nongovernmental parties. Both these elements contributed to a cooperative approach of fiscal policy coordination which differs from Sweden’s rules-driven approach of vertical fiscal relations

47. In the absence of binding fiscal targets for local governments, Denmark has opted for a system of formalized budget cooperation. Budgets for both local and central governments are formulated as part of formal annual budgets, which have been gradually developed since the 1970s (OECD 2002). The agreements are multi-annual and stipulated between local government associations and the central government. They set both aggregate spending limits and tax rates, and also identify main fiscal policy objectives and initiatives in individual policy areas. The format of the agreements has been flexible with varying degrees of details on commitments ranging from declarations of intent to specific targets for particular services. Negotiations are fully aligned with the budget preparation calendar and are instrumental to mustering support in parliament for the approval of the annual block grant to local governments.

48. The intragovernmental agreements in Denmark are however not binding and thus have only limited incentives for compliance. Since the local government association cannot legally commit to fiscal targets for municipalities, enforcement depends on voluntary compliance by municipalities. In the past, the cooperative approach has only been partially successful. Local governments have typically responded to expenditure overruns by raising income taxes and constraining labor supply. This has undermined the credibility and efficiency of the accords between the center and the municipalities. A main shortcoming of the agreements is a lack of incentives for local governments to adhere to the set targets. A particularly interesting case is Austria, which has included financial incentives into its fiscal policy contract between the different levels of government (Box I.1.)

49. In contrast, Sweden’s vertical fiscal cooperation between the different government levels is based on informal consultations, but final policy decisions are made autonomously. While the Spring and the Fall Budget contain medium-term fiscal projections for the general government, they represent the central government forecasts and do not reflect the consensus forecast by all levels of government. There is a steady dialogue between the different government levels on policy relevant issues, but the discourse is not embedded in a formalized coordination process and not intended to arrive at common decision on financial or policy matters. As a result, decisions on the size of grants from the central government, expenditure plans local tax policy are made independently.

50. While Sweden’s rules driven approach is in line with principles of self-government, the law would not preclude a closer cooperation between governments in the budget planning process. A more formalized exchange of information at the technical level could help to better assess rising fiscal demands and pressures on local tax policy. Effective medium-term planning is also hindered by uncertainties about volatile tax revenue and frequent changes to government grants. Finally, there are no additional incentives for municipalities to adhere to fiscal policy targets and to contribute to macroeconomic tax policy objectives.

F. Conclusion

51. Sweden has a long tradition of decentralized public service delivery. Local governments are responsible for more than 40 percent of general government expenditure and absorb more than two thirds of overall public consumption. In international comparison Sweden stands out with greater local responsibilities in demographically sensitive areas and a very high degree of local tax policy autonomy. In addition to the common education and childcare duties, local spending is strongly tilted towards health and long-term care, which is sensitive to demographic changes.

52. In recent years pressures on local public finances have led to rising local income tax rates. Since 1993 the average (population weighted) local tax rate has increased from 29.9 to 31.5 percent in 2004. These increases have offset one quarter of tax relief granted by the central government. The study has examined the role of various possible explanations and has come to the following conclusions:

  • Underfunded mandates have been a main reason for tax rate increases. Rising expenses for long-term and disability care and an aging population have significantly increased local tax rates, while changes in child-care or education expenses have not translated into tax rate hikes.

  • The local tax base has not weakened. Growth of the local tax base and higher grants from the central government have helped eased pressures to raise taxes, but could not avert them.

  • Cyclically induced overspending may have played a part. However available data did not allow to directly test for this possibility.

53. Looking forward, Sweden’s local public finances will face continued pressures from demographic changes. The effects of population aging will be felt somewhat earlier in Sweden than in many other EU countries. While recent reforms have put the public pension system on a sustainable foundation, increases in demand for health related services will strain the fiscal capacities of local public finances. Growth in health related expenditures, primarily due to cost increases, are likely to exceed savings from childcare and education. An important issue will be that the government explores reform options which do not adversely affect labor supply.

54. Balancing local public finances without further increases of the overall tax burden will therefore be key to avoid fiscal sustainability problems in the long-term. Sweden’s tax burden is already high and stands at 50 percent of GDP. Additional local tax rate increases will adversely affect labor supply and could begin to backfire when population aging begins to bite. Diversification of the tax base by transferring stable revenue sources to the local government, may help to smooth aggregate revenue fluctuations, but not address the problem of an increasing tax burden, since the lost revenue at the centre would have to be matched through other increases. Similarly, higher vertical grants would only shift the financing problem to the central government without solving it. Thus containing expenditures through either higher efficiency of local or central public services, or a reduction in the already generous welfare benefits system will need to be considered.

55. An important measure to enhance long-term fiscal planning is to improve fiscal policy coordination between the centre and the local authorities. Effective aggregate tax policy planning in particular would require:

  • improved vertical policy coordination,

  • more predictable local financing sources,

  • and the use of incentives to stick to agreed expenditure targets also at the local level.

Sweden’s reliance on a balanced budget rule has diverted efforts to develop a more integrated approach of intergovernmental fiscal policy planning. Budgets of local and central government are drafted in parallel with limited exchange on policy objectives including tax policy. Effective medium-term planning at the local level is hindered by uncertainties about volatile tax revenue and by frequent modifications of central grants. Experience from Denmark and Austria shows that many of these shortcomings can be addressed through a formalized budget preparation process, in which local and central government reach agreement over medium-term expenditure targets, tax rates, and vertical grants. Making such an arrangement effective in Sweden would involve that transfers to local governments become more formula driven, their plans are embedded in a medium-term setting with observable targets, and their attainment is tied to incentives. An important element would also be high public visibility and transparency of agreed targets to heighten public accountability.

References

  • Aronsson, T, Lundberg J., and Wikstrom M. (2001) Regional Income growth and Net Migration in Sweden 1970-1995, Regional Studies, Vol. 35/9 823-830.

    • Search Google Scholar
    • Export Citation

  • Banca d’Italia (2001) Fiscal Rules. Proceedings of the Research Department Public Finance Workshop on Fiscal Rules, Rome.

  • Daban, T., Detragiache, E. di Bella G., Milesi-Feretti G., and Symansky S. (2003) Rules-Based Fiscal Policy in France, Germany, Italy, and Spain, IMF Occasional Paper, Washington DC.

    • Search Google Scholar
    • Export Citation

  • Hallerberg, Mark, Rolf Strauch, and Jürgen von Hagen. 2001. “The Use and Effectiveness of Budgetary Rules and Norms in EU Member States.” Report Prepared for the Dutch Ministry of Finance, July.

    • Search Google Scholar
    • Export Citation

  • Joumard, I. and P.M. Kongsrud (2003) Fiscal Relations across Government Level, OECD Economic Department Working Paper No. 375.

  • Mueller, D. (2004) Why Federalism?, in. E. Molander (ed): Fiscal Federalism in Unitary States, Kluwer.

  • Mueller, D., and E. Uddhammar (2004) Strengthening Federalism in Sweden, in. E. Molander (ed): Fiscal Federalism in Unitary States, Kluwer.

    • Search Google Scholar
    • Export Citation

  • OECD (2004) Economic Survey–Sweden, Paris.

  • Per Pettersson-Lidbom and Matz Dahlberg (2003) An Empirical Approach for Evaluating Soft Budget Constraints, Mimeo.

  • Soderstrom, L. (2003) Combining local self-government with long-run discipline and shortterm flexibility with reference to Sweden, in Decentralization: Trends, perspectives and issues at the threshold of EU enlargement, Danish Ministry of the Interior and Health 2003.

    • Search Google Scholar
    • Export Citation

  • Ter-Minassian, T. (editor) (1997) Fiscal Federalism in Theory and Practice, International Monetary Fund, Washington DC.

  • Thakur, S., M. Keen. B. Horvath and V. Cerra (2003) Sweden’s Welfare State—Can the Bumblebee Keep Flying? (International Monetary Fund, Washington DC).

    • Search Google Scholar
    • Export Citation

  • Wendorff, K. (2001) The Discussion on a National Stability Pact in Germany, in Fiscal Rules, Proceedings of the Research Department Public Finance Workshop on Fiscal Rules (Banca d’Italia).

    • Search Google Scholar
    • Export Citation
1

Prepared by Stephan Danninger.

2

Defined as municipal plus country council tax rates

3

They are defined in specific legislation such as the Social Services Act, Health and MedicalServices Act, the Environment Code, and the Education Act.

4

The thresholds for the national income tax are SEK 284.300 and SEK 430.000 respectively.

5

One exception was the specific grant for pharmaceuticals until 2003, which covered all costs ofpharmaceuticals in the context of healthcare services provided by the county councils.

6

The year 1993 has been chosen as the first year unaffected by large fluctuations of the exchange rate.

7

The comparison period was determined by the availability of data.

8

Defined as the share of population age 65 and older in working age population

References

  • Allen, S.G. 1981. An Empirical Model of Work Attendance. The Review of Economics and Statistics 63 (1): 77–87.

  • Andersen, T., and P. Molander (eds.) 2003a. Alternatives for Welfare Policy, Cambridge University Press

  • Andersen, T., and P. Molander 2003b. Policy options for reforming the welfare state in Andersen, T., and P. Molander (eds.) 2003a

  • Andrén, D. 2001a. Short-term Absenteeism Due to Sickness: The Swedish Experience, 1986-1991. Working Paper in Economics, No. 46, Department of Economics, Göteborg University.

    • Search Google Scholar
    • Export Citation

  • Andrén, D. 2001b. Long-term Absenteeism Due to Sickness: The Swedish Experience, 1986-1991. Working Paper in Economics, No. 47, Department of Economics, Göteborg University.

    • Search Google Scholar
    • Export Citation

  • Andrén, D. 2003. Sickness-related Absenteeism and Economic Incentives in Sweden: A History of Reforms. CESifo DICE Report 3/2003.

  • Arai, M., and P. Skogman Thoursie. 2001. Incentive and Selection in Cyclical Absenteeism, FIEF Working Paper Series No. 167.

  • Audas, R., and J. Goddard 2001. Absenteeism, seasonality, and the business cycle. Journal of Economics and Business 53 (2001): 405–419.

    • Search Google Scholar
    • Export Citation

  • Brown, S., and J.G. Sessions. 1996. The Economics of Absence: Theory and Evidence. Journal of Economic Surveys 10 (1): 23–53.

  • Barmby, T.A., Ercolani, M.G., and J.G. Treble. 2002. Sickness Absence: An International Comparison. The Economic Journal 112: F315–F331.

    • Search Google Scholar
    • Export Citation

  • Barmby, T.A., Ercolani, M.G., and J.G. Treble. 2004. Sickness Absence in the UK: 1984-2002. Swedish Economic Policy Review 11, forthcoming.

    • Search Google Scholar
    • Export Citation

  • Bergendorff, S., Berggren, S., Cohen Birman, M., Nyberg, K. Palmer, E., Skogman Thoursie, P., and J. Soderberg 2004. Sickness Absence in Europe–A Comparative Study. Social Insurance Studies no. 2, Swedish National Insurance Board, forthcoming.

    • Search Google Scholar
    • Export Citation

  • Bonato, L., and L. Lusinyan 2004. Work Absence in Europe, IMF Working Paper, forthcoming.

  • De Jong, P., and M. Lindeboom 2004. Privatisation of Sickness Insurance: Evidence from the Netherlands. Swedish Economic Policy Review 11, forthcoming.

    • Search Google Scholar
    • Export Citation

  • Drago, R., and M. Wooden. 1992. The Determinants of Labor Absence: Economic Factors and Workgroup Norms Across Countries. Industrial and Labor Relations Review 45 (4): 764–778.

    • Search Google Scholar
    • Export Citation

  • Dunn, L.F., and S. A. Youngblood. 1986. Absenteeism as a Mechanism for Approaching an Optimal Labor Market Equilibrium: An Empirical Study. The Review of Economics and Statistics 68 (4): 668–674.

    • Search Google Scholar
    • Export Citation

  • European Foundation for the Improvement of Living and Working conditions, European Industrial Relations Observatory (EIRO) on line (various issues), http://www.eiro.eurofound.eu.int.

  • Henrekson, M., and M. Persson. 2004. The Effects on Sick Leave of Changes in the Sickness Insurance System. Journal of Labor Economics 22 (1): 87–113.

    • Search Google Scholar
    • Export Citation

  • Ichino, A., and R. T. Riphahn. 2004. The Effect of Employment Protection on Worker Effort. A Comparison of Absenteeism During and After Probation. http://www.iue.it/Personal/Ichino/short44.pdf

    • Search Google Scholar
    • Export Citation

  • International Labor Organization (ILO) 2003. Key Indicators of the Labour Market (KILM), CD-ROM.

  • Johansson, P., and M. Palme 1996. Do Economic Incentives Affect Work Absence? Empirical Evidence Using Swedish Micro Data. Journal of Public Economics 59: 195–218

    • Search Google Scholar
    • Export Citation

  • Johansson, P., and M. Palme 2002. Assessing the Effect of Public Policy on Worker Absenteeism. Journal of Human Resources 37: 381–409

  • Kaivanto, K. 1997. An Alternative Model of Pro-Cyclical Absenteeism. Economics Letters 54: 29–34.

  • Larsson, L. 2002. Sick of being unemployed? Interactions between unemployment and sickness insurance in Sweden, Institute for Labour Market Policy Evaluation (IFAU) Working Paper 2002:6

    • Search Google Scholar
    • Export Citation

  • Larsson, L. 2004. Harmonizing unemployment and sickness insurance: Why (not)? Institute for Labour Market Policy Evaluation (IFAU) Working Paper 2004:8

    • Search Google Scholar
    • Export Citation

  • Leigh, J.P. 1985. The Effects of Unemployment and the Business Cycle on Absenteeism. Journal of Economics and Business 37 (1985): 159–170.

    • Search Google Scholar
    • Export Citation

  • Lindbeck, A. 1977. The Swedish Experiment. Journal of Economic Literature 35 (3), 1273–319

  • Palme, M., and I. Svensson 2003. Pathways to retirement and retirement incentives in Sweden, in Andersen T., and P. Molander (eds.) 2003a.

    • Search Google Scholar
    • Export Citation

  • Scruggs, L. 2004. Comparative Welfare Entitlements Data Set. University of Connecticut.

  • Shapiro and Stiglitz 1984. Equilibrium Unemployment as a Worker Discipline Device. American Economic Review 74 (3), 433–44

  • Skogman Thoursie, P. 2002. Reporting Sick: Are Sporting Events Contagious?, Stockholm University, Department of Economics, Research Paper in Economics 2002:4

    • Search Google Scholar
    • Export Citation

  • US Social Security Administration, Social Security Programs Throughout the World (SSA). (various issues).

  • WEO 2004, World Economic Outlook. International Monetary Fund.

APPENDIX

Table A.1.

List of Countries and Data Availability

article image

Missing data: Germany (1984), Italy (1992), Netherlands (1984, 1986)

Table A.2.

List of Variables, Definitions, and Sources

article image
Table A.3.

Determinants of Sickness Absence–Static Panel Data Model Dependent variable: share of persons absent due to sickness in total employed

article image
t-values in brackets. ** (*,+,-) = significant at 1 (5, 10, 15)-percent level. All regressions include time fixed effects (not reported).

LSDV estimates with robust errors, fixed effects of other countries are not reported.

Table A.4.

Determinants of Sickness Absence—Dynamic Panel Data Model Dependent variable: share of persons absent due to sickness in total employed

article image
t-values in brackets. **(*,+,-) = significant at 1 (5, 10, 15)-percent level. Arellano-Bond GMM estimates are reported. Regressions include time fixed effects (not reported). AR(2) test refers to the test of the null of no second-order autocorrelation in the first- differenced residuals. Regressions pass the Sargan test of overidentifying restrictions.
1

Prepared by Leo Bonato and Lusine Lusinyan. This paper has benefited from comments from ClasOlsson, Gun Alm Stenflo, Krister Andersson, seminar participants at the Swedish Ministry of Finance, the Institute for Labour Market Policy Evaluation in Uppsala, and IMF headquarters. The authors are grateful to the Eurostat NewCronos LFS team for their assistance with the data on working time, and Lyle Scruggs, University of Connecticut, for providing the dataset on the main aaracteristics of social security systems across countries. The authors thank Xavier Debrun for providing the dataset on employment protection and Haiyan Shi for her excellent researa assistance

2

The aart shows an increase in working hours on average over the subperiods, whia masks the constant decline experienced since 1999 (Figure II.3). The data on average hours are intended for comparisons of trends over time and are not suitable for comparisons of levels (OECD Employment Outlook, Statistical annex).

3

See European Industrial Relations Observatory on line, Working Time Developments, (various issues).

4

Palme and Svensson (2003) find that 20.5 percent of males and 26.9 percent of females use sickness benefits as a pathway to retirement, the second most important avenue after state old-age pensions

5

See Ministry of Finance, Budget Bill for 2004, Budget Statement, p.16. (http://www.sweden.gov.se/content/1/c4/37/05/bbaf6e4f.pdf)

6

See, for example, Henrekson and Persson (2004); Andrén (2001a, 2001b, 2003); Johansson, and Palme (1996, 2002); Skogman Thoursie (2002).

7

See Ministry of Finance, 2004 Spring Fiscal Policy Bill, Budget Statement, pp.20–21 http://www.sweden.gov.se/content/1/c6/02/03/01/9f243de3.pdf

8

Brown and Sessions (1996) provide an extensive survey of the theoretical literature on labor absence.

9

The analysis in this section draws on a forthcoming working paper (Bonato and Lusinyan, 2004), to whia we refer for details.

10

Employees are grouped into two main subgroups: those who worked at least one hour during the reference week, and those who had a job, but did not work at all during the reference week. The reasons provided for absence—defined as a positive difference between usual and actual hours of work—are thirteen for the first group and nine for the second group. We refer to sickness absence as that due to own illness, injury, or temporary disability. Sickness absence of those in the first group is defined as ‘short-term,’ that of those in the second group is defined as ‘long-term.’

11

Tables A.3 and A.4 present the results in detail.

12

The Netherlands have taken a radical approaa in 1996, making employers responsible for the full cash benefit payment. Most firms, however, opted to reinsure their sick pay liability with private insurance companies, reducing the incentive effect. Nonetheless, De Jong and Lindeboom (2004) do not find any difference in absence rates of firms that opted for reinsurance. Although any conclusion from that experience is still tentative, absence started declining three years later and has now dropped below the Swedish level.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Sweden: Selected Issues
Author: International Monetary Fund
Volume/Issue:
Volume 2004: Issue 245
Publisher:
International Monetary Fund
ISBN:
9781451835960
ISSN:
1934-7685
Pages:
52
DOI:
https://doi.org/10.5089/9781451835960.002