During the past decades, income inequality has triggered policies aiming at social inclusion in democracies worldwide. The strengthening and deepening of social security and removing barriers to the educational and health system are common ingredients of a welfare policy, to mention some. The spatial dimension of inequality inspired elaborate regional policies at the national and supranational (namely the EU) levels. Yet, persistent regional income disparities have cast severe doubt on regional policies’ effectiveness in alleviating spatial inequality. Critics of the regional policy approach point to countries’ underlying structural characteristics, such as the immobility of labour and capital, agglomeration effects, and weak institutions that hinder a smooth regional convergence.
Recently, international comparative research highlighted the critical role of countries’ institutional setup in dealing with regional disparities, particularly their degree of fiscal decentralisation. Subnational revenue and spending autonomy would provide strong incentives to local and regional governments to boost their economy. If this hypothesis is confirmed, federations and unitary countries with strong subnational governments would be well equipped to deal with regional imbalances compared to highly centralised ones.
In general, transfers or grants from the central government complement autonomously raised state and local tax revenue. These vertical transfers are justified from an equity point of view since subnational governments differ in their tax-raising capacities and face different costs of providing public services. In this case, central government transfers are designed to guarantee in all jurisdictions equal access to essential public services such as education, housing and health. In many federations, horizontal transfers, namely between subnational governments, are organised to equalise to some degree their revenue-raising capacity.
In contrast to the assumed incentive promoting role of own taxes, transfer revenue has received a rather negative connotation in the literature on fiscal federalism since it would weaken the receiving government’s accountability. Federal transfers’ unconditional character could induce subnational governments to allocate them to less productive expenditure, thus hampering convergence.
The purpose of the research summarised here was to empirically investigate the simultaneous impact of subnational tax autonomy and vertical transfers on the regional convergence process.
Measuring regional income disparities
The key variable in the measurement of regional disparities is gross domestic product (GDP) per capita. The data were obtained from the OECD regional database covering 1995-2011 and refers to 30 countries.
A ‘region’ is defined according to the TL2 territorial classification referring to states and provinces, except for the small countries where TL3 (small regions) is used: Belgium, Estonia, Ireland, the Netherlands, New Zealand, the Slovak Republic and Slovenia. Australia is, for example, represented by eight territorial units, Canada by 13, the United States by 51 and Switzerland by 7.
Commonly used disparity measures are the coefficient of variation and the Gini coefficient of regional GDP per head. Whereas the coefficient of variation measures the average deviation from the arithmetic mean, the Gini coefficient shows to what extent the data differ from a hypothetical distribution where all observations are identical, thereby reflecting perfect equality. Therefore, the Gini coefficient has been widely used in studies on income distribution. It has, for comparative purposes, the advantage that it ranges between 0 and 1, pointing to increasing inequality.
Over the period 1995-2013, the Gini coefficient’s regional disparities slowly increased in all OECD countries. Marked differences in inequalities between the OECD countries are reflected in the 1995-2011 average Gini coefficient where Slovenia ranks highest (Gini of 0.36) and Japan and Sweden lowest (Gini of 0.07), preceded by Australia and Switzerland (Gini of 0.09).
Measuring subnational tax autonomy and transfer dependency
Fiscal decentralisation can be approached either from the subnational expenditure or revenue point of view. One would presume that as state and local governments control a larger share of total government expenditure, the country’s fiscal decentralization scores higher. However, such a result is not justified since part of subnational public expenditure, for example, on education and infrastructure, may be mandated by the central government, thus constraining subnational expenditure autonomy. Therefore, subnational revenue autonomy is a preferred measure of fiscal decentralisation.
A good indicator of subnational fiscal autonomy can be derived from the OECD Fiscal Decentralization Database tax autonomy tables. According to their discretionary power over rates and (or) reliefs, these tables distinguish tax revenue for local governments and regions. Autonomous tax revenue, transfers, shared taxes, user fees, property income and loans constitute the primary sources of subnational government revenue.
One would expect that states in federations benefit from a relatively high degree of fiscal autonomy. Indeed, autonomous tax revenue represents about 50% of total revenue in respectively the provinces, cantons and states of Canada, Switzerland and the US. According to the OECD data, subnational governments in Australia and Spain derive about one-third of their revenue from autonomous tax income. Mexico and Germany are situated at the other end of the spectrum with an autonomous tax share smaller than 10%. It is worth noting that from the point of view of fiscal autonomy, local governments in the Scandinavian unitary countries Sweden and Denmark score high (60 and 30% respectively of autonomous tax revenue) in contrast to their European counterparts.
Subnational governments characterised by a low degree of fiscal autonomy, such as in the Belgian and Mexican federation, complement their autonomous tax revenue with transfers or grants from the central government that represent about 70% of their total revenue. Taxes shared with the central government constitute an important revenue source, for example, for the German Länder and local governments in most unitary states. In this case, the tax base and rates and tax reliefs are determined at the central level, whereas a sharing formula defines the total revenue’s subnational distribution.
Factors determining regional disparities: empirical results
As was found in related research, the empirical results confirm that the level of development is an important driver of regional inequality. In the early stages of development, growth tends to be concentrated in centres in which capital and skilled labour fuel the process of growth. As development proceeds, diseconomies of agglomeration and environmental constraints put the growth poles at a disadvantage in favour of the initially lagging regions, thus narrowing disparities. This ‘inverted U’ path of disparities is confirmed by the results and implies a maximum level of GDP per capita ranging from US$40,000 to US$54,300, from which disparities decrease. Since only four countries in the sample, that is, Norway, Sweden, Switzerland and the US, enjoyed over the period studied such income levels, one may conclude that in most of the 30 countries, regional disparities rise at higher levels of development.
Furthermore, the results confirm the virtuous impact of subnational tax autonomy on regional convergence, although its effect weakens as subnational governments are more transfer-dependent. Nevertheless, transfers from the central government are apt to contribute to regional convergence, a redistributive policy target they were designed for. However, their marginal contribution to convergence diminishes as subnational governments become more transfer-dependent. In the same line of thought, political decentralisation was shown to foster convergence.
Finally, it appears that more densely populated countries display smaller regional disparities than their less populated counterparts because of their lower intraregional transaction costs, favouring the spatial diffusion of technical and organisational innovations.
Concluding remarks
Both subnational tax autonomy and vertical transfers were shown to be potential drivers of the regional convergence process.
From the point of view of accountability and incentive value, tax autonomy allows lagging regions to compete with their more developed counterparts successfully and attract investment and households, thus broadening their tax base. Their tax sources vary from the property and local business taxes to surcharges or rebates on the personal income tax, depending on the institutional environment.
Institutional characteristics, such as a high degree of political centralisation, may not favour fiscal decentralisation, as is the case in many unitary countries, for example France. Vertical transfers are then, in principle, apt to contribute to convergence, although their marginal contribution to convergence diminishes and turns negative when sub-national governments become highly transfer dependent.
Further reading
Van Rompuy, P. (2021). Does subnational tax autonomy promote regional convergence? Evidence from OECD countries,1995-2011, Regional Studies, 55(2), 234-244. https://doi.org/10.1080/00343404.2020.1800623.
“Institutional characteristics, such as a high degree of political centralisation, may not favour fiscal decentralisation, as is the case in many unitary countries, for example France. Vertical transfers are then, in principle, apt to contribute to convergence, although their marginal contribution to convergence diminishes and turns negative when sub-national governments become highly transfer dependent.”
Paul sub- national governments don’t have to worry about vertical dependence when a federal government isn’t budget constrained. They are only dependent when the federal government is budget constrained. Sovereign currency issuing governments with a free floating currency, no debt in a foreign currency and who set their own interest rates, can always provide federal fiscal funds to sub- national governments, as long as they know those funds are being utilised efficiently and effectively. Of course sub- national governments should always strive to have as much fiscal independence as possible.