CHALLENGES AND OPTIONS FOR REFORMING THE SSN SYSTEM IN UKRAINE A REFORM PROPOSAL DEVELOPED AS PART OF TECHNICAL ASSISTANCE ASA P168853 TTL: Nithin Umapathi Team: Caterina Ruggeri Laderchi, Roman Zhukovskyi, Katerina Petrina, Mikhail Matytsin 5/7/2019 SUMMARY A REFORM PROPOSAL Despite the significant progress Ukraine made on its structural reform agenda, social assistance reform remains a key priority in the face of high levels of poverty (estimated at 27 percent according to the WB methodology, 2016 data) and continuing fiscal pressures. The modernization of the safety-net can provide much more effective welfare support to the population while consuming fewer fiscal resources. The key to this approach is fiscal consolidation and rebalancing of the fiscal resources from Housing Utilities Subsidy (HUS) to Minimum Income Support (GMI) for improved targeting of social safety net spending. This would entail a gradual contraction of the HUS and, the expansion of the GMI, and the integration of some child benefits such as the single mother benefit into the GMI. Such integration of support is expected to improve targeting of social spending, reduce fiscal pressures, improve benefits for poor households, simplify the system, standardize rules, and discourage welfare dependency. The end-goal for the proposed reform is to develop a system that supports more equitable distribution of welfare expenditures, with the GMI playing a central safety net role, and HUS playing a smaller complementary role to support energy affordability for most vulnerable groups. 1 Contents 1. Introduction .......................................................................................................................................... 3 2. Challenges and opportunities in Ukraine’s Social Assistance system ................................................... 4 2.1. The Guaranteed Minimum Income program ................................................................................ 7 2.2. The Housing and Utility Subsidy program..................................................................................... 8 3. Seizing the opportunity to modernize Ukraine’s safety net ................................................................. 8 3.1. Expanding the GMI as the backbone of the safety net in Ukraine ............................................... 9 3.2. Rationalizing the HUS .................................................................................................................. 10 4. Conclusions: the way forward............................................................................................................. 11 Figures Figure 1: Proportion of GDP spent on safety nets ........................................................................................ 5 Figure 2: Distributional performance of SSN expenditures in Ukraine......................................................... 5 Figure 3: Composition of SSN spending by type of support (percent of GDP) ............................................. 6 Figure 4: The GMI reform impact on the incomes of the most typical GMI family (2 adults, 2 children) .. 10 Figure 5: GMI and HUS integration impact on incomes of most typical GMI family (2 adults, 2 children) .................................................................................................................................................................... 11 2 1. Introduction Despite the significant progress Ukraine made on its structural reform agenda, much remains to be achieved on several fronts, including improving the effectiveness of social assistance. Social assistance reform remains a key priority in the face of high levels of poverty (estimated at 27 percent according to the WB methodology, 2016 data) and continuing fiscal pressures. Improving the effectiveness of social assistance in Ukraine, while involving many of the policy challenges that are common to many countries in the world (such as ensuring that transfers to the poor and vulnerable are designed to provide effective relief from hardship while creating incentives to work for those who can aspire to become self-reliant), also involves addressing a very idiosyncratic feature. This is the interlinkage between the efforts to create a more effective energy sector and those to provide effective and targeted support to poor households. These two policy transitions have become interlinked as tariff reform was a major stumbling block to energy sector reform. The government response to the large increases in gas tariffs introduced in 2015 has been centered on scaling up the pre-existing Housing and Utility Subsidies (HUS) program, a program of in-kind subsidies delivered as discounts on the utility bills, which was modified to introduce a strong element of targeting. Note that this approach was tried after the failure to introduce a new temporary cash transfer program targeted at the bottom 30 percent of the population in 2014. In contrast, by relying on well-known and tested implementation modalities such as those of the HUS the country was able to achieve an unprecedented scale up in the program, managing to cover over 40 percent of the population, while still achieving significant fiscal savings. Yet this oversized HUS program (absorbing by itself around 3 percent of GDP in 2017) led to levels of social assistance spending comparable to those of high-income countries (World Bank 2018). In addition, the expansion of the HUS has meant both disincentives to cut demand or invest in energy efficiency (at least for recipients whose consumption falls below the levels assumed by the formula) and the emergence of disincentives to work for HUS beneficiaries. The large coverage of the program, and the relative ease through which some of the eligibility requirements could be circumvented, also contributed to large resentment among non-recipients. But possibly most fundamentally, the overall safety nets became lopsided after the expansion of the HUS. The program became the largest social assistance program reaching close to half of the country’s households in 2017 (6.5 million), while for example the Guaranteed Minimum Income, a program designed to help households cope with poverty, reached no more than 270 thousand households. In subsequent rounds of reforms, the design of the HUS has been tweaked to reduce the amount of subsidy provided, and to introduce greater accountability and transparency. The social norms (coefficients linked to the characteristics of the house and of the technology and equipment used) on the basis of which the amount of the transfer is determined have been tightened; means of cross-checking reported incomes and assets have been introduced through the construction of integrated databases; and perverse incentives – such as incentives to artificially “split� households to maximize subsidy receipt have been eliminated. As a result, number of households dropped from around 7 million to under 4 million. These efforts to rein in HUS spending can be seen as the logical counterpart to the gradual increases in tariffs which were part of the transition to import parity levels of gas prices envisioned at the beginning of the reform. 3 While the transition has not yet been fully achieved, a new radical set of reforms has been introduced in 2019, with the monetization of the HUS for most beneficiaries. With this reform, what was paid as a discount on the bill is now provided as cash equivalent via a state bank and pension fund. Having the household allocation in cash provides incentives to save on energy. From a system’s point of view monetization also offers the advantage of simplification, as the flow of information is managed by the payment authority. To date players in the energy sector have acted as the delivery system for the subsidy but have also been able to benefit from the lack of transparency on the number of beneficiaries they were effectively serving. Under the new benefit delivery system, the potential for commercial malpractice is lower. While a welcome step, the 2019 reforms have highlighted some of the inherent tensions within the system. This in turn has helped identifying the elements of a broad ranging rationalization agenda for the social assistance system. The opportunities and challenges of this process of rationalization are the main topic of this note, which seeks to present a vision of a more effective and sustainable social assistance system for Ukraine. Central to this vision is the possibility to create effective basic income security by integrating targeted social safety net programs while rationalizing programs that are fiscally unsustainable, not well targeted and distort incentives. More concretely, this would involve the expansion of the Guaranteed Minimum Income program, targeted to the poorest groups, financed from a reduction of the Housing and Utility Subsidy. The next section provides more background details on the current challenges of the system, while section 3 presents a vision for change through some simple scenarios and identifies key entry points for implementation and section 4 concludes by summarizing the pathways for change. 2. Challenges and opportunities in Ukraine’s Social Assistance system Ukraine spends around 5 percent of its GDP (2017) on social assistance – as large as in much richer countries such as Germany and the Netherlands and more than double the average for ECA region (Figure 1). And in terms of performance, Figure 2 illustrates that despite such high rate of spending, the social assistance system overall directs too few resources to the poor. In the face of these budget allocations, poverty, which rose significantly after the strong contraction of the economy in 2014-15, is high at an estimated 27 percent (2016) and most of the poor are not supported adequately. This is a challenge of exclusion as well as of the adequacy of support for those who do receive social assistance. The gaps in coverage are particularly marked for groups such as the working poor. Other groups, such as households with children clearly receive insufficient support if poor despite being covered by at least some program. Child poverty is estimated at 38 percent in rural areas, and 31 percent in urban areas. Many of the design features of the country’s social assistance programs reflect a desire to limit welfare dependency and disincentives to work, but, the existing system also transfers a non-negligible amount of resources to the non-poor (see Figure 2) and such features are not effective at reducing disincentives. 4 Figure 1: Proportion of GDP spent on safety nets 6 5.0 5 4 2.7 3 2.2 1.6 1.9 2 1.4 1 0 Source: State of Safety Nets, World Bank Report, 2018 Figure 2 A and B: Distributional performance of SSN expenditures in Ukraine Ukraine’s social assistance system is characterized by the Generosity of payment for poorest households is low and twin issue of partial coverage of the poor and inclusion of not much higher than payment received by better off better off households households (1 USD = 25 HRV) 90 78 400 80 343 Percent of population covered 350 HRV per month per person 70 64 298 300 276 265 57 259 60 52 250 50 43 200 40 150 30 100 20 10 50 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Quintile of the per capital expenditure distribution Quintile of the per capita expenditure distrubtion Source: Household Budget Survey (2016); World Bank calculations based on HUS, GMI and the Child Birth Grant Despite this disappointing performance, the large budget devoted to social assistance in the country implies that there is a large opportunity to redirect resources and rationalize the system in a way which increases its impact. Just to put the numbers into perspective, the amount spent on social assistance in 2016 was UAH 115 billion, and the total poverty gap1 for 2016 was UAH 46 billion. For comparison, the total resources 1 Poverty gap is amount of budget needed to eliminate the income gap between the poverty line and household income if it were possible to target those resources perfectly to the population below the poverty line. 5 that went to the top 40% of the population were UAH 37 billion, which suggests that improvement in targeting theoretically can redirect resources to substantially improve the poverty situation in the country. Central to this finding is the unbalanced nature of the system (see Figure 3), with the HUS and the (untargeted) child birth grant, accounting for over 60 percent of the overall SSN budget, against 0.4 percent of the GMI. Figure 3: Composition of SSN spending by type of support (percent of GDP) 0.2 0.5 0.4 0.3 1.7 1.7 Disability benefits Old age benefits War veterans and labor veterans Families and children benefits Housing & Utilities Other It is important to stress that the existence of this opportunity for “rebalancing� social assistance expenditure towards measures that are narrowly targeted at poverty reduction does not mean that a social assistance system should be entirely devoted to targeted transfers aimed at filling the difference between an individual’s income and the poverty line. Indeed, as Error! Reference source not found. discusses, a number of different characteristics that go well beyond targeting performance help define the desirable characteristics of a social assistance system, and there is debate around many of them. For example, the objectives pursued by the system might go beyond focusing on the income poor only to encompass those who are at risk of falling into poverty due to a shock; the need to be dynamically efficient might result in program design that does not strictly targets benefits, but rather lets them taper off for the better off; and the desire to avoid errors of exclusion or to foster a sense of social cohesion might result in an aspiration towards universalism rather than targeting, as far as some needs are concerned at least. Similarly, while ideally a consolidated system with common eligibility criteria for all programs might appear desirable, shifting to such a system and even operating it might result in significant losses of support for some of the most vulnerable groups. One area where the opportunity for strengthening the targeted component of the social assistance system in Ukraine should be considered seriously, however, is programs explicitly targeted to the poor and lower middle-class households. In the Ukrainian case this means the Guaranteed Minimum Income (GMI) program and the Housing and Utility Subsidy program (HUS), which jointly account for around 70 percent of the social assistance budget, with HUS consuming most of the share. These two programs are very 6 different in purpose and in their targeting mechanisms, with the GMI intended to give support to households unable to provide subsistence for themselves, while the HUS aims to ensure that a basic bundle of energy consumption is affordable for lower income households.2 Crucially, the two programs are very different in size, with the GMI covering less than 3 percent of households in Ukraine, while the HUS, despite significant efforts to reduce its reach still covers 30 percent of households (4.5 million). Recent changes to the HUS program, which is now going to be disbursed in cash to households as opposed to as a discount in the utility bills, are now bringing to the fore the anomaly of a large subsidy to about 1/3 of the country’s population to address one specific need (energy prices), while the overall needs of certain categories of poor are not appropriately covered by the GMI. 2.1. The Guaranteed Minimum Income program The Guaranteed Minimum Income (GMI), is a tightly targeted program aimed at extremely poor families that need last resort support to achieve the subsistence minimum level of income. It is implemented as a cash payment to households, and the benefit is equal to a difference between the guaranteed level of subsistence minimum (21 percent of the subsistence minimum for able-bodied, 85 percent for children and 100 percent for disabled) and household’s total income. The benefit is also topped up for families with children (UAH 250/month for children aged 0-13, and UAH 500/month for children aged 13-18). The maximum benefit is set at 75 percent of the subsistence minimum for the family. Note that by artificially lowering the amount of income guaranteed (the Guaranteed Subsistence Minimum (GMS) threshold is only a portion of the Subsistence Minimum, for all beneficiaries other than the disabled), this design limits program coverage. The children top-ups, added to soften the impact of the reforms of the Child Benefit program in 2016, concentrate assistance on families with more children. As a result of these design features, the GMI is a relatively small program which falls short of providing appropriate coverage even for the groups identified as poor by the official poverty definition. At the heart of this limited effectiveness is a concern with avoiding welfare dependence, which results in an effort to limit the coverage of able-bodied individuals. Trying to address two objectives – reducing poverty and reducing welfare dependency – with one tool means that the program is not very successful in achieving either of them. Note that the marginal “tax� on additional income that household member can earn by working is 100 percent – meaning that the incentives to work are limited unless recipients can access jobs that pay them significantly more than the guaranteed minimum. Such strategy results in undesired effects such as the limited support offered to households with children if they have able bodied parents, accompanied by high rates of child poverty. In contrast with this practice, international experience (starting with the 1990s “welfare to work� reform of social assistance in the USA) suggests that concerns for welfare dependency can be better addressed by designing a program which offers appropriate protection for those in need, while providing incentives for 2 As Ukraine’s fiscal crisis escalated in 2014-15, the government increased tariffs to close energy-related fiscal imbalances. At the same time, to ensure that these tariff increases were socially and politically sustainable, the small existing Housing and Utility Subsidy (HUS) program – which contributes to the cost of housing-related utilities of its beneficiaries - was scaled-up and redesigned to become more progressive. Enrolment conditions were loosened and take-up was encouraged at scale. The scaling up of the HUS, together with the adoption of a path of tariff increases towards import parity rather than a full tariff shock, were explicitly part of a transition strategy to allow time for households to adapt to the higher tariff environment. 7 self-reliance. Those are typically embedded in a “disregard� formula, that allows prospective workers to keep a share of their benefit, so that by working they experience a net increase in income (i.e. their earnings do not replace “hrvnia by hrvnia� the benefit). 2.2. The Housing and Utility Subsidy program The HUS provides targeted financial support for the payment of utility bills by capping how much households spend to cover a normative amount of consumption. Normative consumption is defined by the social norms – a set of parameters varying by characteristics of the housing unit (e.g. amount of space where heating is needed, type of heat and gas connection, utility type), number of household members, and heating or non-heating season. A key feature of the design of the program is that as incomes increase assistance decreases – in other words, the cap on normative consumption increases with income. This is achieved by having a key parameter – the eligibility threshold, i.e. the maximum share of household income that normative consumption on energy should account for – declining with income. As the threshold is on a sliding scale depending on income, in 2016 the formula resulted in an effective eligibility threshold (i.e. maximum share of out of pocket expenditures on housing and utilities) varying from 4 percent on average for the bottom decile to 15 percent for the reference household (a household of two with per capita income equal to twice the “subsistence level�). Since the share of income spent on utilities declines with income, this means that 92 percent of the households in the bottom income decile were eligible in 2016, compared to 14 percent of households in the top decile. As the program covers the excess cost of normative consumption over 15 percent, someone in the top decile, for example, who on average would have an eligibility threshold of 25 percent, would not be covered by the HUS. As mentioned in section 1, the HUS has been the key mechanism for ensuring that an unprecedented shock in gas tariffs in the transition towards import parity pricing could be socially sustainable. Absorbing nearly 3 percent of GDP and 50 percent of the social assistance budget in 2017, however, the program had grown unsustainably large and has generated large scale unintended consequences. Covering over 40 percent of Ukrainian households at its peak, the program has generated significant resentment among non- beneficiaries, with the growth of perception on egregiously rich recipients in the popular press. The way incomes were calculated has also given incentives for perverse behaviors, such as artificially splitting households or limiting labor demand, not to displace transfers with higher earnings. Possibly equally damaging has been the cushioning of incentives to save energy for recipients able to spend less than normative consumption through energy saving or investing in energy efficiency measures. At the same time, the program despite its size is designed to address only one particular source of vulnerability in household livelihoods. Finally, the program also shows in-built exclusionary biases. To qualify households, need both low income and high bills, thereby excluding poor rural households who might not be connected to consume either district heating or gas.3 3 Households depending on other heating sources, such as coal or wood, could resort to other forms of support, mostly centred on direct fuel supply. Such programs are far less generous than the HUS. 8 3. Seizing the opportunity to modernize Ukraine’s safety net Given this context, there is huge opportunity to address the inherently lopsided nature of Ukrainian social protection system – a system where most of the social assistance budgets is spent on a single cause of poverty and vulnerability, such as higher utility prices, through a program that by design cannot cover all the poor. Rebalancing the relative weight of the HUS and the GMI could result in a system producing net fiscal savings vis-à-vis current expenditure, thanks to stronger targeting of the HUS and the rationalization of the system. At the same time the system could grow in effectiveness, reducing poverty, particularly among children, and producing dynamic benefits thanks to the reduced work disincentives and increased investment in energy saving/energy efficiency. Seizing this opportunity requires addressing forcefully the weaknesses of both the GMI and the HUS, as this section discusses with references to the evaluation of a set of reforms to both programs. 3.1. Expanding the GMI as the backbone of the safety net in Ukraine Our proposals for strengthening the GMI focuses on two key priorities, identified through a detailed analysis of the program’s eligibility rules and available data: (a) expanding the program’s coverage and generosity and (b) ensuring that the program provides incentives to promote self-sufficiency and exit from the program. • Expanding program coverage: The crucial parameter determining coverage (and generosity) of the GMI is the level of income that is guaranteed. This varies across categories of beneficiaries, but other than for disable recipients it is a fraction of the subsistence minimum. Changing those fractions for key groups is the simplest way of ensuring greater coverage of the program – a precondition for the program to become the backbone of modern social assistance in Ukraine. Our recommendation is for two thresholds in particular to be revised: o For working age adults: the threshold should be revised from 21 percent of the subsistence minimum to 33 percent. o For children: the program should guarantee a full subsistence minimum for every child, while doing away with a complicated system of top ups which has proven not enough to result in adequate support. • Strengthening incentives for self-sufficiency: efforts to address the plight of child poverty and the low levels of support for working poor households with children might prove ineffective if more comprehensive support is not coupled with measures that preserve incentives to work. To this effect, and to address the 100 percent marginal tax on any additional income that households earn, we suggest introducing an earned income disregard at the rate of 35 percent on any new or additional incomes. This would result in a gradual phasing out of the benefit as households find alternative sources of income to the benefit. The incentives for self-sufficiency could also be strengthened by penalties to encourage labor activation, such as the loss of part or all GMI benefit if there are able bodied household members reporting less than the minimum wage or no income without being registered with the public employment services. 9 As Figure 4 illustrates the joint implementation of these measures would marginally improve the incomes of a typical GMI recipient family with no income, but would totally transform the income profile of the same family struggling to make a living by earning a minimum wage income. The implementation of these measures would increase coverage from the current 2 percent of population to around 5.5 percent of the population, while on average the benefit per household would not increase. GMI expenditures are expected to increase by 120 percent. However, reducing official poverty overall by 40 percent, child poverty by 70 percent, and poverty among households with working adults by 60 percent seem to make this a much more cost-effective policy deal than continuing to invest in the HUS . Figure 4: The GMI reform impact on the incomes of the most typical GMI family (2 adults, 2 children) 3.2. Rationalizing the HUS Just as the scaling up of the HUS was conceived to help the transition towards a new set of energy prices, its scaling back will sanction the country’s success in adapting to a new set of economic incentives. This second phase of the transition will therefore see the HUS take a more limited role, as a support for energy affordability for those whose affordability challenge is mostly dictated by the (temporary) inability to moderate expenditures with energy efficiency investments than by low incomes (which would be mostly taken care of by the GMI). Complementary investments by the Energy Efficiency Fund, about to become operational, will help families like those still covered by the HUS managing their consumption and eventually graduating from the HUS. While this second phase of the HUS has already started, with progressive reductions of the social norms, including a new round taking place at the end of the 2018/19 heating season, further action will be needed. Our recommendation focuses on: • Improving targeting and reducing coverage through a higher eligibility threshold, set at 20 percent (rather than 15 percent) for the reference household. The sliding nature of such threshold with lower income households facing lower eligibility thresholds should continue. However, it is important to note that a higher eligibility threshold will also create further pressures on household budgets of vulnerable households. To address this risk, it will be necessary to create rational exempt categories such as households with elderly pensioners with modest incomes, and households that include persons with disability. 10 • Greater synchronization of eligibility conditions with other means tested benefits. Such measures alone could be expected to decrease HUS expenditures by nearly 20 percent annually relative to the 2018 baseline, while decreasing coverage by nearly 1 million from the current 4 million. As Figure 5 illustrates despite fiscal savings due to declines in the HUS generosity, current beneficiaries of the GMI would see their incomes protected and even boosted. Figure 5: GMI and HUS integration impact on incomes of a typical GMI family (2 adults, 2 children) Family has 2 Adults, 2 Children - no income Family has 2 Adults, 2 Children – 1 minimum wage 7000 7000 6000 6000 5000 5000 4000 4000 3000 3000 2000 2000 1000 1000 0 0 Current GMI - no income GMI_Reform - no income Current GMI GMI_Reform - no GMI_Reform - EARNED income working poor HUS benefit GMI benefit Total income (including GMI and HUS) (earned income) HUS benefit GMI benefit Total income (including GMI and HUS) 4. Conclusions: the way forward This note has summarized a vision for a more modern social assistance system in Ukraine, one that recognizes the significant achievements of the last few years, while also acknowledging that ongoing fiscal pressures and high poverty call for continuing reforms. The end-goal for the proposed reform is to develop a system that supports more equitable distribution of welfare expenditures, with the GMI playing a central safety net role, and HUS playing a smaller complementary role to support energy affordability for most vulnerable groups. The policy recommendations that have been detailed focus on strengthening and expanding the GMI, while rebalancing fiscal resources by reducing the coverage and size of the HUS program. Further fiscal resources to support the expansion of the GMI could come from the consolidation of small and ineffective programs such as the Single Mother benefit4. Such integration and rebalancing of fiscal resources towards the GMI is expected to improve targeting and adequacy for poor households, reduce exclusion of the poor, simplify the system, standardize rules, while containing fiscal pressures and welfare dependency. Implementing these reforms calls for careful balancing of fiscal savings with expanding the protection afforded to the poorest. In this light, the GMI can be expanded at a faster pace first to prevent adverse 4 This benefit shares eligibility and means-test criteria with the GMI. Therefore, there is a case to consolidate it with the GMI while improving the GMI program. 11 impact on poverty and thus prepare the groundwork for more accelerated contraction of the HUS starting from 2019. Note that some of these changes are mutually reinforcing – for example, because of the way income is accounted for in the HUS, an expansion of the GMI would automatically crowd out some of the HUS transfers, thereby facilitating the HUS reduction. In addition, a greater policy focus on the GMI, would help address the implementation bottlenecks which currently see its budget spent only up to 80 percent of the plan. 12