Chapter 5. Building a social protection system

The MoLSD’s structure should be optimised for a systemic approach

The MoLSD has responsibility for establishing a social protection system in Kyrgyzstan. However, the structure of the central Ministry is not well aligned to a systemic approach, in part because the Ministry’s composition and mandate have changed on numerous occasions since independence, generating frequent changes that have created instability and made it difficult to align the institutional architecture with its functions.

A Ministry of Labour and Social Protection was established in 1992. Since then, the status of the institution has changed. In 2009, for example, the Ministry of Labour was a stand-alone institution while the Social Protection component was downgraded to government agency. In 2015, the Ministry of Labour, Migration and Youth with the Ministry of Social Development were amalgamated to form the MoLSD, with the Minister of Social Development placed in charge of the combined portfolio.

As highlighted by functional analysis of the Ministry of Social Development carried out by USAID in 2014, these changes have strained critical systems related to information technology, human resource management, financial management and monitoring and evaluation (USAID, 2014[1]). At the same time, instability has reduced the extent to which the MoLSD’s structure matches the policies it is mandated to implement. The challenge is particularly acute for the labour portfolio, reducing the capacity to generate synergies with the social protection portfolio that can realise the promotive potential of social protection.

The merger of the Labour and Social Development portfolios is relatively common internationally. However, the structure of the MoLSD consists of a labour department and an employment department rather than units aligned to specific active labour market policies that are able to stimulate the labour market participation of vulnerable groups. (Figure 5.1). Specialised units co-ordinating and devising policies for public employment services, public works programmes, vocational training and microcredits should be considered. The work of these units should be able to call on strong analytical capacity within the Ministry regarding current trends in the labour market, in particular changes in the skills that employers are demanding.

Moreover, the current strategy for labour and employment was designed for implementation by the Ministry for Labour, Migration and Youth rather than the MoLSD. This misalignment is likely to undermine the likelihood of its effective implementation, especially if institutional links to the State Migration Service and the State Agency for Youth Affairs are weakened as a result of the break-up of the former Ministry.

Figure 5.1. The MoLSD’s structure reinforces fragmentation

Organogram of the Ministry of Labour and Social Development (MoLSD)

Note: The numbers within each box designate the number of officials assigned to the aforementioned function or unit.

Source: MoLSD (2016[2]), Order Number 2 of 01 Nov. 2016, Kyrgyzstan.

The structure of the social protection components of the MoLSD better reflects the social protection strategies for 2012-14 and 2015-17, with specific units for the three vulnerable groups prioritised in the latter (families in difficult life situations, people with disabilities and the elderly). However, the structure is not well aligned to a systemic approach. Social protection policy making is divided across two divisions, each headed by a different deputy minister. One division contains units responsible for programmes (state benefits and compensations) and financing, while the other division is structured according to vulnerable groups. Given that there is a strong link between the design of different programmes and the groups they are meant to serve, this separation is not intuitive.

Meanwhile, the specific unit in the Ministry responsible for co-ordination of social protection policies is not overseen by a Deputy Minister but by the Permanent Secretary. It thus risks being excluded from the core policy-making activities and appears to perform an administrative function, alongside units responsible for human resources, legal services and information technology. A co-ordination unit located above the Deputy Ministers that reported directly to the Minister would be better placed to oversee co-ordination and thus the development of a social protection strategy. The unit for policy co-ordination should also be expanded to cover labour and employment.

The new social protection strategy (2018-22) is an opportunity to bring the social protection and labour portfolios closer together. It would be optimal if the structure of the MoLSD were reconfigured to maximise the potential for this to occur.

Looking further ahead, this report proposes that non-contributory arrangements currently managed by the Social Fund be brought under the MoLSD. For this to happen, specific capacity would be required within the Ministry and its structure would need to change accordingly. This is likely to be extremely complex from an administrative perspective; such a transition should be planned and implemented over a long timeframe with careful attention paid to retaining the highly skilled staff of the Social Fund.

Stronger capacity at local level is critical to enhance provision of social protection

Social protection provision is decentralised, with the majority of MoLSD staff employed at rayon level: before the Ministry’s amalgamation in 2016, some 70% of the MoLSD workforce operated at a sub-national level (GoK, 2013[3]). However, responsibility for delivery of services is fragmented between the aiyl okmotu and the MoLSD staff with responsibility for a specific aiyl aimak. The uncertainty around provision at the local level makes it hard for national policies to gain traction at a local level and has created confusion around the MoLSD’s role.

At a central level, the Ministry is responsible for social protection policy making, designing strategies, policies and legislation that correspond to the demands of the executive authority. At the same time, it is also responsible for monitoring the implementation of social protection policy, analysing the performance of sub-national social protection staff and developing policies in response to information that is generated at the local level.

The inter-governmental structure of the MoLSD conforms to a clear systemic logic. Policies are made at the central level but implemented at the local level in collaboration with local government on the basis that local-level workers will inform the centre how well these policies are responding to the needs of individuals, households and communities. Staff at rayon level were initially intended to provide oversight, monitoring and co-ordination of local-level staff and social protection activities.

In practice, the weak capacity of the social workers employed by the aiyl okmotu undermines the effectiveness of social protection policy and requires rayon-level staff to play a more active role at a local level. Not only is local government unable to provide social services at a community level but it also struggles to support the implementation of national policies, in particular the identification of households eligible for state benefits. At the same time, households identified as poor and vulnerable are excluded from the benefits to which they might be eligible and do not receive other forms of social support to promote inclusion and well-being.

Filling this gap in the social protection system will not be straightforward. For the MoLSD to take full responsibility for the provision of social services would require changes not only to the legislation governing social protection but also to the Constitution and budget laws to formally shift responsibility for the provision of social services to the MoLSD. Such a radical change might have significant political costs.

Whichever model is chosen for strengthening local services will also require a major change of focus within the Ministry from an emphasis on payment of social protection benefits to the holistic provision of social protection, including services. The addition of a specific unit for social services to the structure of the MoLSD would be an important first step in this direction.

Such a reform would also require a significantly higher allocation to the MoLSD’s activities at a local level, both to fund the salaries of social workers and give them the resources to fulfil their functions. It would thus raise the Ministry’s salary costs in a context where the GoK is seeking to reduce the public sector wage bill. New financial arrangements that bring together the MoLSD and non-state actors are likely to be essential for providing the resources required.

It is important to note that investment in the local social workforce is likely to enhance the MoLSD’s overall operations. In particular, it would improve the GoK’s capacity to target benefits at the intended beneficiaries, thereby generating efficiency gains in spending on state benefits. Even with this additional capacity, however, there remains an overwhelming case for simplifying the application process and eligibility assessments for potential MBPF beneficiaries.

A stronger social workforce would also facilitate the co-ordination of social protection policy with other aspects of public policy at the local government level, such as health and education services. Coherence of early childhood services (including benefits) should be considered as a high-priority subsystem within the broader social protection system given the high rates of child poverty and long-term benefits of a healthier, better educated workforce in terms of Kyrgyzstan’s future productivity. Such an approach also has great potential to reduce demand for residential institutions.

Stronger mechanisms are needed for collaboration across the social protection sector

In 2016, the MoLSD was made responsible for co-ordinating social protection policy, replacing a council chaired by the Deputy Prime Minister on social affairs that combined representatives from all relevant agencies as well as civil society. However, the MoLSD’s weak influence on social insurance policy, which remains the mandate of the Social Fund, is a fundamental limitation on the development of a social protection system from an institutional, policy, financial, information and citizen perspective.

There is not a clear mechanism for discussions between the MoLSD and the Social Fund and it not apparent that the Social Fund takes the MoLSD into consideration when designing policy. This is not to say that the Social Fund is not cognizant of the challenges it confronts. However, there does not exist a formal process for addressing these challenges that includes the MoLSD; without this, it will not be possible to formulate a systemic policy response.

As discussed elsewhere in this report, non-contributory schemes represent a growing burden on the Social Fund. Moving these programmes to the MoLSD would restore the contributory and non-contributory arrangements and ensure that the required reforms to the pension system take into account non-contributory and contributory arrangements (as well as the labour-market implications of different reform options).

The GoK should consider establishing a task team headed by the Vice Prime Minister to examine long-term reform options for retirement arrangements. As well as the Social Fund and MoLSD, the task team should include the Ministry of Health, the MHIF (which is part of the contributory system), the Ministry of Finance, the State Tax Service as well as the Tri-Partite Commission of government, trade unions and workers and civil society organisations.

Figure 5.2. Responsibility for social protection rests with MoLSD and the Social Fund

Co-ordination of all activities in the social sector

Note: Model protocol of interaction between government agencies of executive branch.

Source: GoK (2013[4]), Government’s Decree No. 404 from 9 July 2013, Kyrgyzstan.

Responsibility for co-ordination of social protection with other areas of social policy falls under the Vice Prime Minister on social issues (Figure 5.2). An example of productive collaboration across the social sector is in the provision of health care, where vulnerable groups are exempt from co-payments for various services and thus partly protected from the impoverishing impact of out-of-pocket health expenditure. However, challenges with the identification of vulnerable households undermine this mechanism.

With the mandatory health insurance scheme and the SGBP providing financial protection against health care payments, formal sector workers and many vulnerable individuals are at least partly covered. However, the resources available to this arrangement are limited: of the 27.25% contribution by formal sector workers, only 2% goes towards the health insurance. The resultant financial constraints mean that insurance beneficiaries receive support with small health-care expenses but not with larger and more substantive outlays, leaving households vulnerable to major health shocks.

At the same time, the expansion of the exemption of co-payments through the SGBP to cover children under age 5, pensioners older than 70 years and pregnant women in 2004 led to a surge in hospitalisations, increasing demand for hospital services. By subsidising health care costs, the social protection interventions in health care substantially increase the financial and institutional load on the health care sector, without resourcing the rising burden.

The shortfall of human resources to manage the demands, inefficient manual management systems, financial unsustainability and a lack of preventive health initiatives are four key supply-side constraints to Kyrgyzstan’s health sector. An important step would be the completion of the Unified Health Information System, which will improve the procurement and management of health care services.

Meanwhile, fragmentation of responsibility for residential institutions is an impediment to the government’s optimisation strategy (discussed in Chapter 2). Although different ministries involved in this sector are responsible for different types of institution – such as those for infants, children or people with disabilities – the MoLSD should be empowered to oversee all such institutions to ensure policy coherence and implement the required information systems as well as monitoring and evaluation. The MoLSD’s capacity to monitor private institutions should also be enhanced to ensure they provide an appropriate level of care and abide by the same standards and protocols as public institutions.

A universal grant for infants will have multiple benefits

Current discussions regarding reform options for the MBPF recognise the drawbacks inherent both to a universal benefit and to poverty-targeting. A hybrid option comprising universal benefits for newborns and infants and poverty-targeted benefits for older children is emerging as a means of balancing affordability with impact on child poverty, but this should be complemented by other interventions under a systemic approach to maximise its impact.

Cost concerns render a hybrid option more feasible than a universal approach. However, this should be combined with reforms to the administration of the MBPF to overcome the targeting challenges identified previously. As mentioned above, strengthening the capacity of local-level social workers and simplifying the process for assessing eligibility are essential, so too are steps to raise the means test threshold to a higher level. Proposals to increase from 30% to 100% the proportion of beneficiary households visited by the MoLSD to verify their eligibility might reduce errors of inclusion but will not address errors of exclusion.

This report recommends initially establishing a universal benefit up to a child’s second birthday as a means of ensuring all children receive some support during the first two years of their lives. This is not only a crucial period in a child’s physical and cognitive development, but it is also a time when children require the most attention, meaning it is difficult for parents to combine care with paid work.

Within five years of implementation, eligibility should be increased to age three, at which point children can be enrolled in pre-school institutions. It is critical that coverage of pre-school institutions, particularly in rural areas, be expanded to ensure this transition for children takes place, both for children who are eligible for the MBPF and for those who are not. The cost of this higher eligibility age will be partly offset by a decrease in the number of children being born over this period, which will reduce demand for the newborn and infant benefits.

The advantages of an infant benefit extend beyond the immediate welfare of beneficiary children. Enrolling children on an information system for state benefits at an early age allows for the early identification of vulnerable children. It would also mean that every individual born after a certain date would be included on such a system, thereby providing social protection planners with detailed knowledge of future populations and removing some of the administrative hurdles associated with, for example, internal migration.

The basic contributory pension should evolve into a universal social pension

The MoLSD’s mandate to design policy for the social protection sector as a whole entitles it to determine the policies of the Social Fund, which is responsible for developing and implementing policies for social insurance, paying benefits and collecting pension contributions. In reality, co-ordination between the two institutions is limited and the Social Fund continues to take the lead in social insurance policy.

This lack of co-ordination between MoLSD and the Social Fund constrain the potential for the systematisation of social assistance and social insurance. This is a major concern in a context where the number of people contributing to the Social Fund (and thus covered by social insurance) is gradually declining and coverage of social assistance is also being reduced, meaning the population as a whole is much less protected, both today and looking ahead. The financial imbalances of the system also require remedying action.

In Kyrgyzstan, as in many countries, programmes exist that cover the same eventuality but with different financing mechanisms (contributory and non-contributory) reflecting individuals’ income level. This is the case with pensions and the MSB, with both systems providing benefits to people with disabilities or the elderly. Challenges arise when the value of the benefits paid by the contributory system approaches the level of the non-contributory benefit, which is only accessible by individuals with no contribution records. This is likely to disincentivise individuals from contributing, undermining the social insurance principle and threatening the system’s sustainability.

This phenomenon is apparent in Kyrgyzstan. Although the value of the social pension component of the MSB is low (and declining in real terms), it is still worth more than the pensions received at retirement for some individuals whose contribution period did not attain the threshold required to receive the full value of the basic pension. The same applies for individuals with incomplete contribution histories who have suffered a disability.

The declining tendency for workers to contribute to the Social Fund (in full or at all) and the reduced value of contributory pensions will reinforce each other; the resulting erosion of the system’s sustainability will be compounded by growth in the elderly population. As individuals opt out of the Social Fund, demand for the MSB will increase. The recent increase in the age of eligibility for the MSB above the statutory retirement age for the contributory system (65 against 63 years for men and 60 against 58 years for women) will have little long-term impact on this pressure but has created an inconsistency in the public provision for old age that impedes coherence across contributory and non-contributory pensions.

At an institutional level, the distinction between the contributory and non-contributory arrangements is becoming increasingly blurred. As discussed in Chapter 4, the quasi-contributory basic pension component is now entirely funded by the Republican Budget, in common with non-contributory pensions for the military and other pensions and pension top-ups or increments.

The sharp increase in spending on military pensions is an issue for public finances as a whole rather than just the Social Fund, and a reform of this arrangement is a priority. Bringing expenditure on military pensions under control is likely to require a combination of higher retirement ages and the introduction of contributions for those eligible for military pensions. It would be very difficult to institute such changes for existing members of the military (and other agencies eligible for military pensions), which heightens the imperative of establishing these rules for new staff at the earliest opportunity.

Aside from military pensions, it might be optimal to bring the tax-financed instruments under the administration of the MoLSD, including the basic pension component. If this were to happen, it would be increasingly difficult to justify maintaining the old age and disability components of the MSB and the basic pension component as different systems given that they are each financed from the state budget and there is no link between the value of an individual’s contributions and the benefit they receive from the basic pension component. Replacing the relevant aspects of the MSB and the basic pension component with a universal social pension would greatly simplify the system, make it fairer and easier to manage financially.

A costing exercise of such a reform is beyond the scope of this paper but the World Bank Public Expenditure Review on pensions (2014[5]) calculates that replacing the current basic pension and the MSB with a universal pension available in the event of disability or at age 65 will be significantly less costly than the current system, especially if the grant is indexed against prices rather than wages. However, it also states that this reform would require parametric change to the insurance components.

So far, there has been resistance to such a reform, on the basis that it would undermine the contributory principle. Existing pensioners might also object, on the grounds that their successors will benefit from the new system for free while they had to contribute over their lifetimes. According to the Kyrgyz Republic Pension System Development Concept Paper, the Government is also considering reforms to target the basic pension component towards individuals with low pension pay-outs. However, this raises equity issues along the income distribution.

It is unrealistic to expect a rapid reform to the pension system given political economy dynamics and the present lack of capacity within the MoLSD to expand its role into this domain. However, a reform of some kind appears to be inevitable and the sooner this is planned the easier it will be to transition gradually to a new system. As discussed earlier, pension reform should be an inclusive process that reflects social insurance’s role within a broader social protection system, which it is MoLSD’s responsibility to co-ordinate.

Expand public works programmes for promotive social protection

The low coverage, low financing and minimal impact of active labour market policies are major constraints on the promotive impact of social protection and restricts its ability to break the inter-generational transmission of poverty. The MoLSD should consider scaling up public works programmes as the most direct means of supporting individuals excluded from the labour market. Local government should be at the forefront of identifying appropriate projects but central government would need to be far more involved in the sector.

As identified in Chapter 2, public works programmes are the largest active labour-market policy in terms of coverage. However, they only cover a very small proportion of the unemployed population, in part due to restrictive eligibility criteria. Women are almost entirely excluded. Moreover, there is no national strategy for public works programmes, and little monitoring or evaluation. As a result, public works programmes have very little impact on poor or vulnerable groups.

Relatively simple measures could be adopted to improve this situation. The first step would be to catalogue and analyse the projects that are currently being implemented to understand the current provision of public works programmes. The next would be to design a national strategy for public works programmes that establishes an appropriate balance between the three principal objectives of such interventions: the income support they provide to participants, the training and experience that participants receive and the assets (physical or otherwise) they generate.

In the context of Kyrgyzstan’s broader social protection provision, it is arguable that the second objective is the most important of these (though the first and third should not be overlooked). The emphasis on enhancing the skills of workers means that the programmes should contain a training component and the work should be aligned to the needs of the local labour market. Preferential access (perhaps through quotas) might be established for groups at highest risk of being out of work, particularly women, youth and people with disabilities. The time commitment should balance the need to acquire skills with that of looking for other work; participants might be expected to stay on a project for a number of months but only work two days a week during this period.

To maximise the acquisition of productive skills, public works projects should involve skilled and/or semi-skilled activities. Projects should not be limited to infrastructure development (though this might be very useful in rural areas). Social and environmental projects have become an integral component of South Africa’s Expanded Public Works Programme, for example (Philip, 2013[6]).

Projects should not only serve the skills requirements of participants but also the needs of local communities. Aiyl okmotu could play a role in the selection of participants and the design of projects so that they are of benefit to the broader community. This improves the likelihood that any assets or initiatives generated by the public works project will be maintained once the scheme ends. It would also give aiyl okmotu an important stake in social protection policy in a situation where provision of social services were to become more centralised or involves more non-state actors.

Systematisation can leverage ongoing improvements in the information infrastructure

An important aspect of systematisation is that individuals should have access to social protection programmes appropriate for their circumstances across the lifecycle. To achieve this requires not only that these programmes be in place but also that social protection provision be underpinned by one or more information systems that automatically link individuals to the programmes they need.

Registration for social assistance follows an on-demand approach in Kyrgyzstan. To access benefits, potentially eligible households must complete an application with supporting documents to prove their eligibility to the social protection department. Upon receipt, local-level social workers verify eligibility, which involves substantial time, resources and expertise given the complex income indicators and the remote locations where many potential beneficiaries reside.

The on-demand system requires that individuals know whether they are eligible for a specific programme and can provide the required supporting documents – factors that cannot be taken for granted, especially in very remote areas. The lengthy eligibility assessment also raises the potential for corruption in the application process.

Since 2000, aiyl okmotus have been responsible for monitoring household incomes in their villages, working with the MoLSD to assess eligibility for state benefits and maintaining the SPPF. The GoK has used this SPPF as a management information system with the objective of improving targeting and monitoring and informing planning.

Despite broad coverage, the manual processing and analysis of the SPPF data has created information distortions that have had an adverse impact on targeting effectiveness and generated misleading information for policy making. In response, the government has been developing a Corporate Information System for Social Assistance (CISSA) over a number of years. This is intended to automate all business processes related to acceptance, processing, assignment/rejection and payment of benefits that fall under the responsibility of MoLSD and its departments.

At the time of writing, CISSA includes beneficiaries of the MBPF and other state benefits, the MSB, the Supplementary Monthly Social Benefit and Veterans of the Second World War. It also contains individual modules for specific vulnerable groups, such as people with disabilities (including medical examination data), children or elderly individuals in residential institutions run by MoLSD and recipients of compensations for privileges. It also includes a humanitarian aid database.

CISSA has the potential to revolutionise social protection provision in Kyrgyzstan. It could significantly improve efficiency, accuracy and transparency of data collection, decision-making and policy implementation through automation of these processes. It can also keep track of internal migrants as they move location, allowing for the swift transfer of entitlement from one place to the next. Because the state benefits system is demand-based, the CISSA will contain information on both potential and actual social protection, enhancing its capacity to respond to short-term shocks or evolve to longer-term trends.

However, there are challenges with CISSA that need to be addressed to ensure its viability and sustainability. For instance, there are problems with its operability: the system is reported to be very slow, in part because of the slow internet connections in parts of the country and in part because of the size of the documents that must be uploaded. Reducing the number of documents to be uploaded might ease this problem, though it will be harder to fix if the challenge relates to the country’s IT infrastructure. While these issues are being addressed, the paper-based SPPF remains operational, leading to duplication of effort and gaps or inconsistencies in data collection.

The MoLSD’s capacity to maintain CISSA is also a cause for concern. At present, the system is supported with the assistance of the World Bank. To take over responsibility, the MoLSD will need to ensure that capacity exists at both the national and the district level to maintain its operation. The MoLSD’s IT unit has shrunk over time and it struggles to attract highly skilled IT professionals due to a disparity between public-sector salaries and wages for equivalent work in the private sector.

Another important step in the development of CISSA is to link it to other official databases, thereby creating a single registry for the social protection system that can also be linked to other administrative data sources. In the social protection sector, CISSA should be linked to the Social Fund (especially if pension arrangements become a combination of a universal social pension and the contributory schemes) and the MHIF, each of which operate their own database. It should also be connected to a new database for residential institutions, which is at an advanced stage of development.

It is also envisaged that CISSA will be linked to the State Registration Service (SRS), which gathers a range of administrative data on births, deaths, real estate and vehicle ownership. Since August 2016, the SRS has been responsible for issuing Personal Identification Numbers (PIN) to newborn children. This number was previously issued by the Social Fund upon an individual joining the formal labour force for the purpose of tax and social security deductions as well as accessing health care. Linking CISSA to the PIN system would theoretically ensure that individuals are registered with social protection system at the earliest opportunity.

The SRS has also made considerable progress in developing an electronic unified population register; only individuals included on this register were allowed to vote in the presidential election in 2017. Coverage of the new register is not yet universal, with some individuals unwilling or unable to go through biometric registration. Nonetheless, when complete and integrated with the other databases, the new electronic register will provide information on the entire population and will thus be a major asset for social protection provision and planning. The SRS is presently rolling out information kiosks nationwide to facilitate the registration of marriages, births and deaths as well as applications for ID cards or passports.

The next step is to enhance monitoring and evaluation of social protection. This is an important component of the Programme for the Transition of the Kyrgyz Republic to Sustainable Development 2013-2017, which emphasises the importance of performance management, the implementation of monitoring systems and the evaluation of policies and programmes across government. The programme proposes a mixed-methods approach with a “formative” evaluation in 2016 and a “summative” evaluation in 2017. The results of this evaluation will inform reforms that improve the policies and aim to maximise impact for target groups. The guidelines indicate an M&E plan as well as a staff training plan for those involved in implementation and M&E.

The results of this plan for the social protection sector are not yet available. As a result, analysis of the impact of different programmes (as discussed in Chapter 3) is largely reliant on the KIHS, the national household survey. The KIHS is a valuable tool for social protection analysis, containing specific questions about whether households benefit from different social protection programmes (MBPF, MSB, compensations for privileges and pensions). It is a partial panel survey, which means it is able to follow certain individuals or households over time.

At the same time, however, it is difficult to reconcile the KIHS to administrative data. Part of the complication lies in the proxy means test applied to test eligibility for the MBPF, which is difficult to pick up through household surveys. The varying level of the MBPF benefit (until its standardisation in 2015) and the fact that households might have been eligible for more than one child-related benefit are also likely to have complicated reporting on benefit receipt.

An important aspect of the CISSA will be the extent to which its data serves as a basis for analysis within the MoLSD. This in turn depends on two factors: the usability of the database for analysis and the demand for data within the Ministry. On this second point, the development of analytical capacity within the Ministry has not been a priority owing to the fact that the SPPF has been difficult to analyse. Once the CISSA is fully rolled out, this analytical capacity is essential. At the same time, it would be a rich source of information for external researchers, provided the data could be anonymised.

Social protection reforms must overcome political economy challenges

The political economy around social protection in Kyrgyzstan makes the substantial reforms envisaged in this report problematic. Despite the limited impact of social assistance on poverty and the financing shortfalls for social services and labour market policies in particular, it will be difficult to increase the social protection budget overall both due to a lack of popular support and because of external pressures. Reprioritisation between programmes is possible but this will create losers as well as winners; recent experience shows that those who lose out will let their discontent be known.

The ongoing discussions around reform to the state benefits are a useful case study. The MoLSD’s plan to universalise the MBPF was opposed by the IMF and World Bank, which considered the reform to be unaffordable given fiscal constraints. The IMF places an emphasis on improving the efficiency of social assistance spending through improved targeting and cautions against the universalisation of the MBPF on the grounds that it might “give disincentives for women to participate in the labour force” (IMF, 2018[7]), a concern that does not appear to have a strong empirical basis.

The reform also experienced opposition within Kyrgyzstan. Upon learning that they would be worse-off under the universal system (Bengard, 2017[8]), a small number of MBPF beneficiaries in Naryn region took to the streets in protest at the reform and went on hunger strike in December 2017 (Podolskaya, 2017[9]). The MoLSD postponed implementation of the reform shortly afterwards. An interim state benefit system will be in place for 2018 while the MoLSD redrafts the reforms (with input from the World Bank) and consults with parliament.

It is not clear whether external or domestic influences played a greater role in persuading the MoLSD to reconsider the reform. However, it is apparent that there was not strong support from other constituencies in the GoK or elsewhere for universalisation of the MBPF. Part of the reason might be that social assistance is a less established component of the social protection system. It might also reflect the fact that the main beneficiaries of such a reform would be individuals who are of working age (such as those identified as the largest poor cluster within the LCA in Chapter 1) and who thus are perceived as being responsible for supporting themselves financially.

The relative lack of prominence of social assistance in the discourse around social protection is likely to be a major impediment to a reform of the pension system that introduces a social pension in place of the current basic pension component. Despite diminishing coverage among the working-age population, social insurance remains an important public institution, founded on the principle that individuals earn an income in retirement through their contributions.

However, the contributory principle is already being substantially weakened, not only by the substantial transfers to the pension system (and the health system) but also by the diminishing replacement rates provided by pensions (and strains on the quality of health coverage). A reform of the basic pension component would be a critical step in institutionalising the transition from the Soviet system to a social protection system appropriate for a modern-day market economy.

Such a reform will not be easy. Pensioners represent a very strong (and large) constituency who are quick to mobilise in response to problems with pension payments or perceived threats to the value of their pensions. This is one factor behind the regular (and relatively generous) indexing of pension payments to wage growth in the economy. It is likely that a reform that aligned the basic contributory pension with the social pension for elderly individuals would be received with great suspicion, both by today’s pensioners and by individuals who have contributed to the Social Fund but not yet retired.

However, the Social Fund’s long-term solvency issues and the inevitable decline in the value of pensions make reform essential. Whatever approach the GoK adopts, it will be important to consult broadly on the proposed reforms, reassure existing pensioners that the value of their entitlements will not be affected and explain that reform is intended to safeguard the contributory system rather than dismantle it. An information campaign regarding the state of the pension system would be an important means of sensitising the population as a whole.

Another significant challenge with changing the design of the bottom pillar relates to the strength of the institutions involved. The Social Fund has emerged much stronger politically than the MoLSD over the past 25 years, not least thanks to the power that pensioners wield. The MoLSD is also more vulnerable than the Social Fund to the demands of the electoral cycle, having to implement policies on behalf of the executive which might be motivated by political rather than technical considerations.

The MoLSD is likely to receive support from the World Bank and IMF in the case of pension reform. The World Bank was instrumental in retaining the basic pension component as part of the 1997 reforms (Andrews et al., 2006[10]) and recently advocated for the creation of a demogrant (World Bank, 2014[5]). Whether the World Bank would (publicly) support a transfer of the non-contributory programmes from the Social Fund to the MoLSD is not clear. It remains lukewarm on the FFDC component, which at 2% will not generate very high benefits for contributors.

The other large-scale reform involves the establishment of legislation governing the provision of social services, including the minimum level of social service provision that should be guaranteed by the state. This process might lead to important changes in how responsibility for provision of social services is apportioned between local government and the MoLSD, which in turn would require changes to the Kyrgyzstan’s political, administrative and financial structures.

While the scale of the crisis confronting social services might justify fundamental reform, greater centralisation of social protection provision risks reducing the participation of local government in the sector and thus the system’s capacity to understand local needs and vulnerabilities. At a larger scale, centralisation risks undermining the building of the Kyrgyz state. For generations during USSR rule, decision-making was highly centralised outside its borders. Establishing and supporting the development of local government structures is a key component of Kyrgyzstan’s emergence as a representative democracy (Gjørven, 2013[11]); great care should be taken when undertaking reforms that might weaken or subvert this process.