3. Twenty years of tax autonomy across levels of government: Measurement and applications
The Network on Fiscal Relations has been assessing the degree of sub-central government tax autonomy in OECD countries for almost two decades. This chapter provides an overview of the methodology used to characterise tax autonomy. After summarising the widespread use of the tax autonomy results by researchers addressing a range of policy issues, the chapter highlights recent trends in sub-central government revenues and presents the results of the latest survey of tax autonomy, completed in 2020. Using the OECD’s tax autonomy methodology, the chapter then assesses local government tax autonomy in the 50 US states. The analysis reveals that US local governments have somewhat more tax autonomy than local governments in the average OECD country. The chapter also includes refinements of the tax autonomy methodology.
Much of the economic and political benefit of decentralised public finance comes from the ability of subnational or sub-central governments (SCGs) to make their own decisions about taxation. A local or regional government that is able to define its own tax bases, tax rates, and other characteristics of a tax has a high degree of tax autonomy or taxing power. To provide accurate cross-national comparisons of the importance of state and local governments in countries’ fiscal systems, it is important to be able to characterise state and local tax systems by their degree of tax autonomy.
Starting in 1995, the OECD began to assess the tax autonomy of state or regional and local governments in OECD member countries. A taxonomy was developed to assess the degree of tax autonomy in each country. Each tax instrument used by state or local governments in a country is assigned one of twelve possible policy-based codes to indicate the extent of tax autonomy for the instrument. The results of this exercise are summarised by calculating the share of total government revenue by level of government assigned to each tax autonomy code. From the inception of the OECD Network on Fiscal Relations in 2003, the OECD has completed the tax autonomy study once every three years, with the latest analysis carried out in 2020, based on data for 2018, the most recent final data available at the time. The results of these tax autonomy studies are disseminated in the OECD’s Fiscal Decentralisation database.1
These tax autonomy indicators provide the following major insights:
A composite measure of tax autonomy shows that local governments in the OECD have significantly less tax autonomy than is suggested by simple expenditure or revenue-based measures of decentralisation. The largest degree of SCG discretion over taxes relates to that for the recurrent taxation of immovable property.
While the profile of SCG tax autonomy differs markedly between federal and unitary countries (Figure 3.1), cross-country differences are most marked for unitary countries. There has been a gradual increase in tax autonomy of SCGs over time.
In a unique application of the methodology to all 50 US states, local governments are found to have somewhat more tax autonomy than in the average OECD country, and rely much more heavily on property taxation.
A substantial number of studies have made use of the OECD tax autonomy indicators to examine the effects of increased SCG decentralisation on outcomes such as long-term economic growth, for which the effects are mixed.
Potential methodological tweaks to address measurement challenges are proposed, along with ways to enhance the timeliness and country coverage of the indicator.
Source: OECD Fiscal Decentralisation database.
The chapter provides an broad overview of the methodology used by the OECD to characterise tax autonomy (Dougherty, Harding and Reschovsky, 2019[1]). The second section introduces the importance of measuring tax autonomy for sub-central governments. The third section explains what indicators are used to measure the degree of autonomy. The fourth section documents how the tax autonomy results have been used by researchers to explore the effect of tax autonomy on a range of policy outcomes. The fifth section presents the main results of the most recently completed tax autonomy study. Finally, the sixth section extends the methodology to measure local government tax autonomy in the United States.
Starting with the seminal work of Wallace Oates (1972[2]), economists have highlighted a number of benefits accruing from a fiscal organisation that provides local governments with a substantial amount of freedom to make their own decisions about spending and taxation. On the other hand, there is also a subsequent literature that identifies difficulties with mobile factors of production, equity concerns and capacity constraints. These arguments suggest that some types of revenues are better collected at the local level as a result of their less mobile bases.
The problems of measuring fiscal decentralisation and comparing the degree of decentralisation across countries are complex and widely discussed. The most frequently used measure of fiscal decentralisation is the share of a country’s total tax revenue raised by its SCGs, or the share of total public expenditure attributable to SCGs. There exists, however, a substantial literature that has argued that these fiscal share measures provide highly imperfect measures of fiscal decentralisation (Owens and Panella, 1991[3]; Ebel and Yilmaz, 2003[4]; Stegarescu, 2005[5]; Bird, 2011[6]; Baskaran and Feld, 2012[7]; Blöchliger, 2015[8]; OECD, 2019[9]). The basic problem is that true fiscal decentralisation requires that SCGs be fiscally autonomous. This means that they are free to decide how much revenue to raise and how to spend their available revenues. Without these capabilities, the efficiency and accountability benefits of decentralisation would be diminished.
Revenue statistics, whether from the OECD’s databases or from country-specific budget documents, indicate the level of government to which revenues are attributed, but provide no indication of whether those governments have the power to define the tax base, the tax rates, or any tax reliefs. Without these taxing powers, the tax revenues, whether from shared taxes or own-source taxes, are functionally equivalent to intergovernmental transfers from a higher-level government (Blöchliger and King, 2007[10]).
One of the primary goals of the OECD’s measure of tax autonomy is to supplement existing revenue statistics by allowing analysts to assess SCG revenues in terms of the tax autonomy of the governments raising that revenue. Combining data on tax autonomy with statistics on the share of tax revenue raised by local government provides a comprehensive picture of the overall taxing power of local governments. By repeating the analysis of tax autonomy on a regular basis, it is possible to explore trends in the taxing power of SCGs, and to identify countries that are enhancing fiscal decentralisation by removing restrictions on subnational taxing power.
A complete picture of the fiscal autonomy of sub-central governments also requires an assessment of expenditure assignments among levels of government and an evaluation of the spending autonomy of sub-central governments. Over the past decade, the OECD has initiated both conceptual and empirical research on the spending power of sub-central governments (Bach, Blöchliger and Wallau, 2009[11]; OECD/KIPF, 2016[12]; Dougherty and Phillips, 2019[13]; Kantorowicz and van Grieken, 2019[14]). This line of research is ongoing within the Fiscal Network, and is discussed further in Chapter 4 of this volume.
The first effort by the OECD to measure the tax autonomy of SCGs used data for the year 1995 from a survey of 19 OECD countries (OECD, 1999[15]). The taxonomy that the OECD developed to assess the degree of tax autonomy was modified when the analysis was updated using data from 2002 and the sample became larger (Blöchliger and King, 2007[10]). The modified taxonomy (discussed below in Section “The use of the OECD’s tax autonomy analysis” has been used in all the updates since then.2
The OECD taxonomy is displayed in Table 3.1. The characterisation of tax systems in terms of tax autonomy is inherently complex. Within any given country, there are many tax attributes, and numerous institutional and administrative details that help define the taxing power of SCGs. In developing the taxing power taxonomy, the OECD has tried to capture the essence of tax autonomy in a handful of indicator codes.
The codes listed in Table 3.1 are arranged in decreasing order of tax autonomy. The “a” codes characterise taxes for which SCGs can determine tax revenue by setting tax rates and defining other attributes of the tax, such as exemptions and credits that influence the amount of tax revenue generated by the tax. The “b” codes are assigned in cases where higher level governments control tax attributes, such as the definition of tax bases and tax credits, but state and local governments have complete, or partial freedom to set tax rates. The “c” code applies when SCGs have no control over tax bases or rates, but are given freedom to set tax credits, exemptions, or abatements, collectively referred to as tax reliefs. The c code is infrequently used.
The “d” codes are used for various types of tax sharing schemes. Under a tax-sharing scheme, tax revenue is levied and collected by a higher-level government, and a specified share of the revenue collected is shared with SCGs. Under the definition of tax sharing used as part of the OECD tax autonomy taxonomy, the determination of this share may have an explicitly equalising function though it is not required to.3
The “e” code is for taxes over which SCGs have no autonomy. The “f” code is only used when none of the other codes are appropriate. Fortunately, the f code is rarely used.
Codes are assigned to tax revenues based on the survey responses of OECD member countries, with the most recent survey having been conducted in 2020. The information for each country is summarised by calculating the share of total tax revenue by level of government (state or local) that is assigned to each tax autonomy code (see Table 3.3).
Since 1999, scholars have used the results of the OECD tax autonomy analysis in analytic studies of various aspects of SCG finance. The basic argument made by these papers is that in assessing fiscal decentralisation, it is very important to consider the extent to which SCGs have control over their own tax instruments. Table 3.2 summarises the most relevant literature about the relationship between fiscal decentralisation and economic growth, highlighting the use of OECD tax autonomy data.
Trends in sub-central government tax revenues
Sub-central governments raise a significant share of total tax revenues in OECD countries, in both unitary and federal systems and have also been relatively stable across time as a share of GDP (Figure 3.2), in contrast with central government revenues, which show greater fluctuation across time.
In OECD countries with a federal structure,4 approximately one-quarter of total tax revenues were received by subnational governments between 1995 and 2018. Approximately two-thirds of these were received by state governments and one-third by local governments. Another quarter of revenues were received by social security funds, with the remaining half of total revenues was received by central governments. Across the periods, local revenues have remained steady at between 2.3 and 2.5% of GDP, while state revenues have increased from 4.5% of GDP to 5.5% of GDP (Figure 3.2, left-hand panel).
The share of subnational revenue in OECD countries with a unitary structure is smaller, on average, than in federal countries: local government revenues amount to between 3.6% and 4.1% of GDP across the period, showing a slow and relatively steady increase across the period. Federal taxes accounted for 21% of GDP.
Source: OECD Revenue Statistics.
These averages conceal much inter-country variation within each group. Among OECD countries with a federal structure, subnational governments received between 0.9% and 16.5% of GDP (Mexico and Canada, respectively) in 2018 (Figure 3.3). Among unitary countries, local governments received revenues ranging from less than 1% of GDP in countries like Estonia and Greece for example to over 10% of GDP in Iceland, Denmark and Sweden. Those last two countries raise the highest amount of revenue from subnational governments (15.4% and 12% of GDP, respectively).
Predominant sources of subnational revenues in OECD countries include income taxation (both personal and corporate), payroll taxes, property taxes, and taxes on goods and services, as shown in Figure 3.4. In 2018, subnational governments in federal countries raised the largest share of their revenues from income taxes (Figure 3.4, left-hand panel) and taxes on property formed the largest source of local government revenues. In unitary countries, local governments received the greatest share of their revenues from income taxes (2.4% of GDP in 2018, on average). In neither group of countries do subnational governments receive social security contributions, which are almost exclusively collected by social security funds.
Figure 3.5 shows a breakdown of subnational tax revenues for each country in 2018, divided into revenues received by state governments (upper panel, federal countries only) and local governments (lower panel). Taxes on income, profits and capital gains were the most significant source of state revenues in five countries even if taxes revenues from property taxes and taxes on goods and services can be considerable. At the local level, taxes on property formed the largest share of subnational revenues in half of the OECD in 2018. Income taxes were the predominant source of local government revenues in 12 countries.
Source: OECD Revenue Statistics.
Taxing power of state and local governments in the 2020 exercise
The most recent tax autonomy survey was conducted in 2020. The results of the survey are presented in Table 3.3 and are based on final revenue data for 2018.5 The first two columns provide data on state and local government revenue as a percentage of GDP and as a percentage of total government tax revenue, respectively.
On average, state or regional governments have a higher degree of tax autonomy than local governments.
Of the 38 OECD countries listed in Table 3.3, only local governments in Australia, Mexico and New Zealand have full tax autonomy (97% or more of tax revenue classified as a1 or a2). However, local governments in an additional 18 countries have a substantial amount of tax autonomy, with at least 80% of their tax revenue classified as either b1, b2 or b3. In contrast, local governments in both Austria and Israel have very limited tax autonomy. Local governments in five other countries rely on shared tax revenue over which they have no control (d3 or d4) for over half of their local tax revenue.
As mentioned above, the share of total tax revenue that is raised by SCGs has been criticised as a measure of fiscal decentralisation because it takes no account of the tax autonomy of SCGs. Blöchliger and King (2007[10]) propose as a “composite indicator of fiscal autonomy” the product of the share of SCG revenue that is considered as autonomous and the SCG share of total tax revenue. In their paper, Blöchliger and King define autonomous taxes as those over which SCGs have “discretion over rates and reliefs.” They operationalise this definition as tax revenue coded as a1, a2, b1, b2 or b3. Because tax revenues coded as b2 are subject to some restrictions imposed by higher-level governments, we refer to the Blöchliger-King definition as partial tax autonomy.
Figure 3.7 shows a scatter plot of Blöchliger and King’s composite indicator of fiscal autonomy applied to local governments using the 2018 tax autonomy data presented in Table 3.3. Countries are ordered on the horizontal axis according to their share of total tax revenue raised by local governments. The vertical axis represents the value of each country’s partial tax autonomy composite index. Countries that are displayed below the diagonal axis (not shown) have limited tax autonomy. For example, in Latvia, where 18.1% of total tax revenue is raised by local governments, but only 13.6% of that amount is considered partially autonomous, the composite index has a value of 2.5% (13.6% of 18.1%).
Note: Local taxes that are coded a1, a2, b1, b2 or b3 are considered partially autonomous (see Table 3.3). For legibility, some country labels have been suppressed.
Source: Own calculations.
Figure 3.8 is based on fully autonomous local government taxes, which are defined as taxes coded as a1, a2, or b1 that represent the highest amount of tax discretion for subnational governments. As on average, 41% of local government taxes in OECD countries are classified as b2, its exclusion from the definition of tax autonomy is the reason why many fewer countries are on or close to the diagonal and many more countries have a composite index of tax autonomy equal to zero. Note also that three Nordic countries, Denmark, Finland and Sweden, raise a very high proportion of total taxes through local government taxation and at the same time retain full autonomy over their local tax systems.
Note: Local taxes that are coded a1, a2, or b1 are considered fully autonomous (see Table 3.3). For legibility, some country labels have been suppressed.
Source: Own calculations.
Changes in tax autonomy 2000 to 2018
Changes in the tax autonomy of SCGs in OECD countries can occur for two reasons. Because some taxes are much more sensitive to economic activity than other taxes, over the course of a typical business cycle the mix of SCG taxes is likely to change. Income tax revenues tend to vary substantially between periods of economic strength and economic weakness, while property tax revenues tend to be much more stable over the course of a business cycle. Policy changes and decentralisation reforms can also affect tax autonomy. SCG tax autonomy can change over time if higher-level governments impose new limitations on the ability of lower-level governments to change rates or relief, or if higher-level governments choose to remove existing limits.
Information on the tax autonomy of SCGs in OECD countries is available on a consistent basis from 2000 onwards. Figure 3.9 summarises the results of the five OECD tax autonomy analyses conducted between 2002 and 2018. The data are presented separately for federal and unitary countries.
There has been little change in the tax autonomy of federal countries between 2000 and 2018. In 2000, local governments in the unitary countries had on average the ability to set the rates (but not reliefs) for 49% of subnational tax revenues. By 2018, this share had risen to 68% of subnational revenues. This increase is attributable in part to a decline in the ability to set tax reliefs for local government and the decline of tax not corresponding to the five tax categories.
Because of the heterogeneity that characterises the US fiscal system, assessing local government tax autonomy requires a separate analysis of each state’s tax system. This section reports on the results of a recent analysis of US local government tax autonomy. Additional methodological details and results can be found in Reschovsky (2019[28]).
The only detailed and comprehensive source of data on the tax revenue of local governments is the annual State and Local Government Finances series.6 Those data, however, are provided on a fiscal year rather than the calendar year basis used for the OECD revenue statistics. Conversion from fiscal to calendar year is complicated by the fact that there exists no consistent definition of local government fiscal years and that some local governments used fiscal years ending in every month of the year.
To construct the calendar year 2014 dataset used in this paper, requires the use of Census Bureau data from fiscal years 2014, 2015, and 2016. The conversion requires three steps. First, for each tax in each state, the share of tax revenue collected using each definition of fiscal year found in that state is calculated. The second step involves combining revenue data from different Census fiscal years, with the way these data are combined depending upon the definitions of fiscal years used by local governments. A graphical representation of this process can be found in Appendix A of Reschovsky (2019[28]). The final step involves taking a weighted average of the local fiscal year-specific revenue estimates calculated in step two using as weights the revenue shares calculated in step 1. The result is a calendar year 2014 tax revenue estimate for each local government tax used in each state.7
As shown in Table 3.4, in calendar year 2014 the tax revenue of US local governments totalled USD 645 billion.8 The table lists each tax used, its OECD tax classification code, and the amount of revenue from each tax and the number of states in which local governments used each tax.
The property tax is used in all 50 states plus the District of Columbia. At least some selective sales taxes (sometimes referred to as excise taxes) and license taxes are used by local governments in most states. The use of other taxes is less widespread, with general sales tax used in 35 states, the individual income tax in 14 states, and the corporate income tax in only 8 states. Figure 3.10 uses 2014 national totals to illustrate the share of local government tax revenue coming from each tax. The figure clearly shows the dominant role played by the property tax.
Source: Calculations from U.S. Census Bureau, Annual State and Local Government Finances, fiscal years 2014, 2015 and 2016.
Measuring local government tax autonomy
The central task in determining US local government tax autonomy is to apply the appropriate tax autonomy code9 to each local government tax utilised in each state. One complication in assigning the tax autonomy codes is that within a single state, local government autonomy with regards to any given tax may differ among local governments. Our approach to this within-state heterogeneity was to assign the tax autonomy code that reflected the dominant situation (in terms of revenue).
A further complication arises with respect to the treatment of local government consumptions taxes, specifically, general and specific sales taxes. With a few exceptions, when used by local governments these taxes are “local option” taxes. This means that a state legislature authorises (provides permission) for local governments to levy a tax.
The question arises about how to classify a local consumption tax that while formally a local option tax is in fact utilised by all, or nearly all, local governments within a state. For the study of US local government tax autonomy, the decision was made to classify a local tax that is levied by 90% or more of local governments within a state at a state-mandated rate as an “e”, meaning that local governments lack autonomy with regard to that tax.10 In cases where all or nearly all local governments levy a local option tax, but at various rates, the tax was classified as “b1” or “b2” to reflect their limited tax autonomy.
After each tax in each state has been assigned a tax autonomy code, the share of tax revenue in each state associated with each code is tabulated, and then the state-specific results are summed over the 50 states and the District of Columbia. The results of these calculations are shown in the column labeled United States in Table 3.5. The right-hand column of the table shows the 2014 unweighted average among all OECD member countries other than the United States of the shares of local government tax revenue assigned to each tax autonomy code.
Local government tax autonomy: The United States compared with the OECD average
The results in Table 3.5 indicate that US local governments have somewhat more tax autonomy than local governments in the average OECD country. About 18% of local government tax revenue in the United States comes from taxes with the highest level of tax autonomy (codes a1 and a2) while it is 13% for the OECD countries. On the other hand, only 1% of tax revenue in the United States is derived from taxes over which local governments have no control. In the average OECD country limits to taxing power are much more common, with 11.5% of revenue coming through tax sharing arrangements imposed by central governments, and 7.8% from taxes imposed on local governments.
Tax autonomy by type of tax
To better understand the US results, Table 3.6 displays the tax autonomy results separately for each tax. The taxes are organised using the OECD tax classification scheme. Because nearly three-quarters of total local government tax revenue in the United States is raised through the property tax, the taxing power associated with the property tax dominates the overall taxing power results. The data indicate that over 90% of property tax revenues are subject to some type of state government-imposed tax rate or tax revenue restriction. This in turn explains why the largest share of overall local government tax revenue is classified as b2 or b3, representing restrictions on rates or revenues. However, local governments have a considerable degree of taxing power with respect to several other taxes, such as license taxes (5200) and specific sales taxes (5100).
By recalculating tax autonomy under the assumption that the US local governments raised tax revenue using the same mix of taxes as used by the average OECD country, local governments in the United States would have a substantially higher degree of tax autonomy than the average OECD country (Table 3.7), with nearly half of local government tax revenue now characterised as having the highest degree of tax autonomy (codes a1 or a2). Clearly, the main factor that reduces US tax autonomy is the widespread presence of limitations placed on the property tax, and the much heavier than average reliance on property taxation (73.9% in the United States compared to the OECD average of 44.7%).
Tax autonomy by state
With a few exceptions, US local governments cannot use taxes that have not been explicitly authorised through state legislation.11 Generally, once a tax has been authorised, individual local governments are free to decide whether to levy the tax. Some taxes are authorised only for certain types of local governments. In other cases, taxes can be authorised only for local governments that meet some criteria, usually defined by minimum population size. The only local government tax that is utilised by almost all local governments is the property tax. Given the tax-specific results shown in Table 3.6, the degree of local government tax autonomy in each state will depend in part on the mix of taxes used in each state.
Figure 3.11 presents a map of the United States, with states divided into four categories indicating the type of major taxes general purpose municipal governments are authorised to use.
Source: Dougherty et al. (2019[1]).
For states in which local governments are authorised to use multiple taxes, the reliance on property tax revenue depends both on the number of local governments actually using alternative taxes, and on the revenue raised from each of those taxes. Based on calendar year 2014 data, Figure 3.12 illustrates the percentage of local government tax revenue in each state and the District of Columbia coming from the property tax. It is evident that the importance of the property tax varies tremendously among states. Only 32% of the tax revenue raised by the District of Columbia comes from the property tax.. At the other extreme, in 13 states, local governments raise more than 90% of their tax revenue from the property tax.
Source: Own calculations using data from U.S. Census Bureau, Annual Survey of State and Local Government Finances, fiscal years 2014, 2015 and 2016.
A myriad of state-specific policies that explicitly limit taxing power of their local governments plays a substantial role in explaining across-state variations in taxing power. Table 3.8 illustrates these differences in tax autonomy in the 50 states and the District of Columbia. The first column of data in Table 3.8 shows, for each state, local government tax revenue as a percentage of state and local tax revenue combined. The remaining columns of Table 3.8 illustrate the wide variation in local government tax autonomy among US states. Overall, no clear regional pattern emerges, although Southern states tend to restrict the taxing power of their local governments more than many other states.
References
[4] Alm, J. and J. Martinez-Vazquez (eds.) (2003), “On the Measurement and Impact of Fiscal Decentralization”, Public Finance in Developing and Transition Countries: Essays in Honour of Richard M. Bird, Edward Elgar Press, Cheltenham, pp. 101-120.
[11] Bach, S., H. Blöchliger and D. Wallau (2009), “The Spending Power of Sub-Central Governments: A Pilot Study”, OECD Economics Department Working Papers, No. 705, OECD Publishing, Paris, https://dx.doi.org/10.1787/223123781022.
[21] Bartolini, D., S. Stossberg and H. Blöchliger (2016), “Fiscal Decentralisation and Regional Disparities”, OECD Economics Department Working Papers, No. 1330, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jlpq7v3j237-en.
[25] Baskaran, T. (2012), “Tax Decentralization and Public Deficits in OECD Countries”, The Journal of Federalism, Vol. 42/4, pp. 688–707.
[7] Baskaran, T. and L. Feld (2012), “Fiscal Decentralization and Economic Growth in OECD Countries: Is There a Relationship?”, Public Finance Review, Vol. 41/4, pp. 421-445, https://doi.org/10.1177/1091142112463726.
[6] Bird, R. (2011), “Are There Trends in Local Finance? A Cautionary Note on Comparative Studies and Normative Models of Local Government Finance”, IMFG Papers on Municipal Finance and Governance, No. 1, Institute of Municipal Finance and Governance, Munk School of Global Affairs, University of Toronto.
[8] Blöchliger, H. (2015), “The challenge of measuring fiscal decentralization”, in Handbook of Multilevel Finance, Edward Elgar, https://doi.org/10.4337/9780857932297.00032.
[22] Blöchliger, H., D. Bartolini and S. Stossberg (2016), “Does Fiscal Decentralisation Foster Regional Convergence?”, OECD Economic Policy Papers, No. 17, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jlr3c1vcqmr-en.
[20] Blöchliger, H., B. Égert and K. Bonesmo Fredriksen (2013), “Fiscal Federalism and its Impact on Economic Activity, Public Investment and the Performance of Educational Systems”, OECD Economics Department Working Papers, No. 1051, OECD Publishing, Paris, https://dx.doi.org/10.1787/5k4695840w7b-en.
[10] Blöchliger, H. and D. King (2007), “Less than you thought: The fiscal autonomy of sub-central governments”, OECD Economic Studies, Vol. 43, https://dx.doi.org/10.1787/eco_studies-v2006-art12-en.
[1] Dougherty, S., M. Harding and A. Reschovsky (2019), “Twenty years of tax autonomy across levels of government: Measurement and applications”, OECD Working Papers on Fiscal Federalism, No. 29, OECD Publishing, Paris, https://dx.doi.org/10.1787/ca7ebc02-en.
[13] Dougherty, S. and L. Phillips (2019), “The spending power of sub-national decision makers across five policy sectors”, OECD Working Papers on Fiscal Federalism, No. 25, OECD Publishing, Paris, https://dx.doi.org/10.1787/8955021f-en.
[14] Kantorowicz, J. and B. van Grieken (2019), “Spending Autonomy of Sub-central Governments: Conceptualisation and Measurement”, OECD Fiscal Federalism Background Papers, https://oe.cd/KvJ19.
[19] Kim, J. and S. Dougherty (eds.) (2018), Fiscal Decentralisation and Inclusive Growth, OECD Fiscal Federalism Studies, OECD Publishing, Paris/Korea Institute of Public Finance, Seoul, https://dx.doi.org/10.1787/9789264302488-en.
[18] Kim, J., J. Lotz and N. Mau (eds.) (2015), Interaction Between Local Expenditure Responsbilities and Local Tax Policy, The Copenhagen Workshop 2013, The Korea Institute of Public Finance and the Danish Ministry for Economic Affairs and the Interior.
[27] Lastra-Anadón, C. and S. Mukherjee (2019), “Cross-country evidence on the impact of decentralisation and school autonomy on educational performance”, OECD Working Papers on Fiscal Federalism, No. 26, OECD Publishing, https://dx.doi.org/10.1787/c3d9b314-en.
[29] Lincoln Institute of Land Policy (2018), Significant Features of the Property Tax, http://datatoolkits.lincolninst.edu/subcenters/significant-features-property-tax.
[17] Meloche, J., F. Vaillancourt and S. Yilmaz (2004), “Decentralization or Fiscal Autonomy? What Does Really Matter? Effects on Growth and Public Sector Size in European Transition Countries”, World Bank Policy Research Working Papers, No. 3254.
[2] Oates, W. (1972), Fiscal Federalism, Harcourt, Brace, Jovanovich, New York.
[9] OECD (2019), Making Decentralisation Work: A Handbook for Policy-Makers, OECD Multi-level Governance Studies, OECD Publishing, https://dx.doi.org/10.1787/g2g9faa7-en.
[15] OECD (1999), Taxing Powers of State and Local Government, OECD Tax Policy Studies, No. 1, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264174030-en.
[12] OECD/KIPF (2016), Fiscal Federalism 2016: Making Decentralisation Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264254053-en.
[3] Owens, J. and G. Panella (1991), Local Government: An International Perspective, North-Holland, Amsterdam.
[28] Reschovsky, A. (2019), “The Tax Autonomy of Local Governments in the United States”, Lincoln Institute of Land Policy Working Paper, No. WP19AR1.
[5] Stegarescu, D. (2005), “Public Sector Decentralisation: Measurement Concepts and Recent International Trends”, Fiscal Studies, Vol. 26/3, pp. 301–333.
[23] Stossberg, S. and H. Blöchliger (2017), “Fiscal Decentralisation and Income Inequality”, Journal of Economics and Statistics, Vol. 237/3, pp. 225-273.
[16] Thornton, J. (2007), “Fiscal Decentralization and Economic Growth Reconsidered”, Journal of Urban Economics, Vol. 61/1, pp. 64-70.
[30] US Census Bureau (2019), Annual Survey of State and Local Government Finances, http://www.census.gov/programs-surveys/gov-finances.html.
[26] Van Rompuy, P. (2016), “Sub-national Tax Autonomy and Deficits: Empirical Results for 27 OECD Countries”, Regional Studies, Vol. 50/7, pp. 1248-1259.
[24] Vermeulen, W. (2018), “Decentralised funding and inequality in education”, in Fiscal Decentralisation and Inclusive Growth, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264302488-7-en.
Notes
← 1. The OECD Fiscal Decentralisation database is available at http://oe.cd/FDdb.
← 2. For the 2014 survey, three “c” codes were combined into a single “c” code.
← 3. With an equalising tax sharing system, the share of total shared tax revenues allocated to SCGs with a low level of resources is increased, while the share going to SCGs with a high level of resources is reduced.
← 4. Ten OECD countries have a federal structure: Australia, Austria, Belgium, Brazil, Canada, Germany, Mexico, Spain, Switzerland and the United States. All other OECD countries are treated as unitary: Chile, the Czech Republic, Denmark, Estonia, Finland, France, Greece, Hungary, Iceland, Ireland, Israel, Italy, Japan, Korea, Latvia, Luxembourg, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Slovenia, South Africa, Sweden, Turkey and the United Kingdom.
← 5. Summary tables reporting the results of the six tax autonomy analyses conducted between 2000 and 2018 are included in the OECD Fiscal Decentralisation database.
← 6. These data are available for downloading at www.census.gov/programs-surveys/gov-finances.html. The dataset provides detailed local government revenue and expenditure data for the sum of all local governments in each state and for the District of Columbia.
← 7. These revenue estimates are not perfect, as there is no way to account for uneven patterns of revenue within a fiscal year.
← 8. On average, revenue from taxes account for 65% of the total revenue local governments raise from their own sources, i.e. excluding intergovernmental grants. However, the share of own-source revenues from taxes varies from 44% to 87% across the 50 states and the District of Columbia (US Census Bureau, 2019[30]).
← 9. The detailed information on individual taxes that provided the basis for assigning the tax autonomy codes came from a multitude of sources. For the property tax, information came from Significant Features of the Property Tax, a website constructed and maintained by the Lincoln Institute of Land Policy (2018[29]). The most frequent source of this information came from state government websites associated with state departments of revenue, or other state government agencies.
← 10. Similarly, in states that set a maximum rate and where all or nearly all local governments utilise that maximum rate, the local tax is classified as “e” indicating no local government autonomy.
← 11. Most states adhere to Dillon’s Rule, a legal principle that limits the authority of local governments. Even in non-Dillon’s Rule states, the authority of local governments to establish new taxes is usually quite limited.