Chapter 5. Assessment of sustainability (Module 4)
This chapter presents tools to analyse the long term financial sustainability of the social protection system. It gives guidelines for a dual analysis that takes into account both the expenditure and revenue side, focusing on the spending dynamics of social protection sector and its constituent programmes for the former and on resources available and how these are generated for the latter. The last part of the module suggests fiscal incidence analysis to combine the revenue and expenditure sides, and to calculate the impact of the existing system of taxes and transfers on equality and poverty.
Analytical dimensions
Module 4 assesses how social protection is financed, answering 4 critical questions:
-
1. Are resources are allocated appropriately across the sector?
-
2. Are social protection programmes sustainable over the long term?
-
3. Does potential exist to expand existing schemes or introduce new ones?
-
4. Are the mechanisms used to finance social protection spending consistent with the objectives of the programmes they are financing?
This analysis is based on a whole-of-government approach. It recognises that social protection is just one area of public spending, competing with many other priorities a government must finance. It also reflects the fact that many sources of financing for social protection also fund other areas of expenditure. Social protection is not viewed in isolation but in the context of a government’s overall fiscal framework. Expenditure and revenues are given equal prominence.
On the expenditure side, this module analyses the spending dynamics of the social protection sector as a whole and of constituent programmes. Through detailed trend analysis, it identifies programmes likely to require greater resources in the future, which can be financed through reprioritisation, either from another social protection programme or from another area of government spending. This analysis incorporates information about the effectiveness of various programmes to ensure optimal allocation of resources across the sector. From a system perspective, this analysis also identifies potential economies of scale that can be achieved through greater administrative or institutional coherence.
On the revenue side, the module examines not only the quantum of resources available to the government but also how these resources are generated, since this can have an important bearing on both the sustainability of this financing and the overall effectiveness of social protection spending. It examines the level, composition and trends of tax revenues and other sources of finance, such as social security contributions, natural resource revenues or official development assistance (ODA). It also examines the sustainability of the current structure of public finances, with reference to the fiscal balance, debt levels and the composition of debt.
These revenues are assessed for their suitability as instruments for financing social protection. For example, many of the new social protection programmes that have emerged in recent years are reliant on ODA, but this source of funding is not appropriate over the long term, since it can fluctuate and donors will look to reduce assistance as countries transition to higher income groups. Similarly, revenues from natural resources can be highly volatile and are often based on finite resources; they thus represent an unstable (and often pro-cyclical) source of financing for programmes that require steady, long-term and (often) counter-cyclical financing.
The module also analyses whether the taxes on which the government relies to finance spending support the objectives of social protection, specifically a reduction of poverty and inequality. If progressive public spending is financed through a regressive tax system, then the overall distributional effect is neutral. Likewise, if higher-income earners accrue greater benefits from a social protection intervention – as is often the case with subsidies, for example – then such spending is not pro-poor and should be reallocated. By calculating the overall impact of taxes and transfers, the module not only provides guidance for tax policy today but also informs the debate about how future revenues required by a social protection system might be raised.
Indicators and data sources
This module relies heavily on administrative data, most of which come from the Ministry of Finance and serve as input to the budget process. Such data might be publicly available online, although programme-level data are not, in which case it is necessary to contact the line ministries responsible. Spending and financing data are often presented as a percentage of gross domestic product (GDP); for the purposes of this module, total public expenditure is a more insightful denominator for spending, while GDP remains a key benchmark for macro-indicators, such as total spending, total revenues or debt measures. Household survey data are required for fiscal incidence analysis.
Methodology
This module analyses social protection spending on various levels, starting with the functional level, which establishes broad spending categories for various activities. Social protection spending is calculated as a proportion of total spending and compared with other spending areas to demonstrate its importance to public spending and identify how its spending trends compare with those of other areas of expenditure. Figure 5.1 shows spending by function in Kyrgyzstan between 2005 and 2015; social protection accounts for almost 30% of public spending, more than spending on health and education combined.
Source: NSC (2016[1]), Government budget expenditures, http://stat.kg/en (accessed June 2017).
The module analyses the economic classification of spending to identify how much the government spends, for example, on transfers, capital projects or civil servant wages. Governments need to achieve a balance among various types of spending – in particular between short-term (current) spending and longer-term public investment – which can influence a government’s long-term finances and the development of the economy.
The module then maps spending on the programmes that comprise social protection. These might not be included within the social protection function group, either because they are not part of the main budget (as is sometimes the case with social security arrangements), they are linked to other areas of spending (such as public works programmes), or they are implemented by donors or non-governmental organisations (NGOs). Social protection spending at a subnational government level might also not be included. After this mapping, aggregate spending on these programmes is generated and its dynamics analysed.
Figure 5.2 shows social protection spending in Kyrgyzstan both in real terms and as a proportion of GDP over 2011-15 by the largest components. It confirms that social protection spending grew by both measures, driven largely by pension payments (including military pensions).
KGS = Kyrgyz Som.
Source: NSC (2016[1]), Government budget expenditures, http://stat.kg/en (accessed June 2017).
The module then analyses the spending dynamics of individual programmes. This multi-level approach assesses the long-term sustainability of the various schemes and the extent to which the system can be expanded in response to the shortcomings or future demands identified elsewhere in the Social Protection System Review (SPSR).
This mapping of social protection expenditure is then overlaid with a mapping of financing. It includes both non-contributory (financed by general revenues) and contributory schemes (funded by individuals, usually workers, and employers) and can reveal how the financing flows can blur these distinctions (Figure 5.3). Kyrgyzstan’s contributory pension system is heavily subsidised by tax revenues, which finance the basic pension component, as well as pension top-ups and military pensions. These subsidies account for considerably more than total spending on social assistance.
Source: NSC (2016[1]), Government budget expenditures, http://stat.kg/en (accessed June 2017).
Various revenue sources are then analysed, including tax (disaggregated by instrument) and non-tax revenues (such as natural resource royalties or ODA). The trajectory of these revenues indicates the robustness of a government’s long-term finances, while broader macroeconomic indicators, such as the fiscal balance and the debt-to-GDP ratio, are analysed to indicate a government’s short- and long-term manoeuvrability. Also analysed are the strength of the revenue-collection system, including compliance rates and tax buoyancy, and the degree of decentralisation.
Figure 5.4 shows ODA flows to Ethiopia between 2007 and 2016 as a percentage of GDP. Ethiopia’s social protection system, in particular the Productive Safety Net Programme, has relied heavily on support from development partners. This support is equivalent to an ever-smaller proportion of GDP, in part reflecting the growth of the Ethiopian economy over the period. However, Ethiopia’s tax revenues have not increased as a percentage of GDP, meaning that public resources have not filled the gap. Ethiopia’s National Social Protection Strategy envisages continued decline in donor support for social protection and highlights the importance of planning for a social protection system financed solely from domestic sources.
Last, the module combines revenue and expenditure analyses through a fiscal incidence analysis that calculates the distributional impact of the existing system of taxes and transfers, as well as its effect on poverty. Figure 5.5 shows how the combined impact of Kyrgyzstan’s extensive system of taxes and transfers is close to neutral: the poverty rate is as high when the population neither pays taxes nor receives benefits as when both are in place, although this does not include in-kind transfers, such as public health or education services.
Source: OECD (2016[2]), “Spending from members from the Development Assistance Committee”, OECD.Stat (database), https://stats.oecd.org (accessed June 2017).
Source: Authors’ calculations, based on NSC (2015[3]), Kyrgyz Integrated Household Survey (database), http://stat.kg/en (accessed June 2017).
This final analysis provides crucial guidance in developing recommendations for the most effective or appropriate revenue or expenditure instruments to address inequality or reduce poverty. It builds on methodologies employed in Organisation for Economic Co-operation and Development publications – Social Cohesion Policy Review of Viet Nam (2014[4]); Divided We Stand (2011[5]); and Growing Unequal? (2008[6]) – as well as the methodology devised by the Commitment to Equity Institute (2017[7]). While the analysis in this chapter relies primarily on administrative data, fiscal incidence analysis combines administrative data with survey information on household or individual income and expenditure.
References
[7] Nora Lustig (ed.) (2017), Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty, Brookings Institution Press and Commitment to Equity (CEQ) Institute, New Orleans.
[1] NSC (2016), Government Budget Expenditures, Finance Statistics Webpage, Bishkek, http://stat.kg/en (accessed on 01 June 2017).
[3] NSC (2015), Kyrgyz Integrated Household Survey, National Statistics Committee of the Kyrgyz Republic, Bishkek, http://stat.kg/en (accessed on 01 June 2017).
[2] OECD (2016), OECD.Stat, OECD Publishing, http://data.oecd.org (accessed on 01 June 2017).
[4] OECD (2014), Social Cohesion Policy Review of Viet Nam, Development Centre Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264196155-en.
[5] OECD (2011), Divided We Stand: Why Inequality Keeps Rising - OECD, OECD Publishing, Paris, http://www.oecd.org/els/soc/dividedwestandwhyinequalitykeepsrising.htm (accessed on 06 December 2018).
[6] OECD (2008), Growing Unequal? Income Distribution and Poverty in OECD Countries - OECD, OECD Publishing, Paris, http://www.oecd.org/els/soc/growingunequalincomedistributionandpovertyinoecdcountries.htm (accessed on 06 December 2018).