Policy instrument choice under globalization: Do authoritarian states choose differently?

Do authoritarian states differ from non-authoritarian ones in their social policy choices? The paper presents new data on childcare deinstitutionalization policies in 15 ex-Soviet countries in Eastern Europe and Central Asia. The data suggest signiﬁcant convergence among countries in the adoption of both dein-stitutionalization policy ‘ends’ and ‘means’, despite drastic differences in political regimes. In particular, I trace the adoption of several instruments, which have been actively promoted by international organizations (e.g., foster care, case management, downsizing of institutions, etc.). The data suggest that author-itarian states in the ex-Soviet region do not differ from non-authoritarian governments in their deinstitutionalization instrument choices, suggesting that political regime is not always a major policy determinant. The results are signiﬁcant because they show that authoritarian regimes can also select modern, non-coercive instruments for childcare deinstitutionalization, which are underpinned by the idea of agentic actors and the centrality of the individual. (cid:1) 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
Do authoritarian and non-authoritarian regimes differ in their social policy choices?This paper investigates the choice of policy 'ends' and 'means' by both authoritarian and non-authoritarian countries, focusing on the policy of deinstitutionalization of alternative childcare in 15 countries -all previously republics of the Soviet Union.
Deinstitutionalization of alternative childcare (hereafter DI) is a policy which aims to ensure that every child grows up in a familylike setting, rather than in a children's home or other types of institutional care (Cantwell, Davidson, Elsley, Milligan, & Quinn, 2012).DI policy has been adopted by a wide range of countries (e.g., Babington, 2015;Huseynli, 2018;Kuuse & Toros, 2019;Ulybina, 2020), particularly triggered by the United Nations Convention on the Rights of the Child (UN CRC) (1989), which recognised the right of the child 'to grow up in a family environment' (UNHR, 1989: 1).The spread of DI policies has been actively promoted by international organizations.The United Nations International Children's Fund (UNICEF) has been central to many DI initiatives around the world, promoting and monitoring the implementation of DI commitments by countries, including in the post-Soviet region (e.g.UNICEF, 2009UNICEF, , 2012UNICEF, , 2013).UNICEF's DI mission in the region is clearly captured in a 1995 paper published by UNICEF's research arm Innocenti: '<UNICEF > must assure that no babies and infants suffocate from deprivation on cold iron cots in sterile institutions.It must drive a firm wedge between civility and barbarism and allow no one to cross it' (Burke, 1995: v).In addition to UNICEF, an increasing number of other transnational actors supported the DI transition in the ex-Soviet region, e.g. the World Bank (Fox & Gotestam, 2003;Tobis, 2000), the European Commission, the European Expert Group on the Transition from Institutional to Community-based Care, and a number of pro-DI advocacy networks and NGOs (Eurochild, Better Care Network, Save the Children, LUMOS, SOS Children's Villages International, Hope and Homes for Children, and many others) (e.g.Kuuse & Toros, 2019;Ulybina, 2020).In other words, post-Soviet countries have been under a lot of normative pressure, and sometimes political or financial (e.g., Ulybina, 2020) pressure, from international organizations to adopt DI policies.These advocacy campaigns stressed that child institutionalization is (a) a violation of human rights, (b) detrimental to children's physical, emotional and cognitive development, and (c) costly compared to community-based options (Csaky, 2009;UNICEF, 2012).
In conditions of intensive international pressure, we should expect countries to adopt DI policies, at least at the high level of problem perception and policy goal commitments.In other words, we should expect countries to adopt the concept of DI policy and promise to reduce their recourse to institutional care.This would be in line with many studies, showing the impact of international forces on national agendas (e.g.see in Howlett & Ramesh, 2003) and high-level policy commitments (Hafner-Burton, Tsutsui, & Meyer, 2008).Less obvious, however, is what policy instruments the governments choose, given significant variation in countries' institutional and political systems, capabilities, and past preferences for policy instruments.
Analytically, based on (Hall, 1993;Howlett & Ramesh, 2003), this paper differentiates between two broad policy levels: policy 'ends' (including the perception of the problem and policy goals) and policy 'means' (i.e. policy instruments, or measures used to attain the chosen policy goals).To explore the cross-national adoption of DI policy, two respective sets of data were collected, relating to: (1) the perception of DI as a distinct policy issue, and DI policy goals, and (2) several internationally-promoted DI policy instruments.These data are used to explore whether countries with different political systems make similar DI policy choices.In particular, I focus on whether authoritarian states, as potentially having a preference for coercive, illiberal instruments, fail to adopt modern, individual agency-promoting measures.
The paper proceeds as follows.Section (1) sets this study in context, by relating it to some relevant studies in world society theory and policy instrument choice.Section (2) discusses the promotion of DI policy by international actors.Section (3) explains how the data were collected.Section (4) presents the new data and discusses it in light of existing literature.

Policy choice as a cultural and political issue
2.1.World society theory: Do countries converge and why is this important?
A large number of studies suggest that in conditions of globalization, states tend to converge in their policy choices.One particular strand of this scholarship is based on the neo-institutionalist world society theory, which highlights that identities and behaviour of national actors are exogenously constructed by the world culture -the culture of the world society (Meyer, Boli, Thomas, & Ramirez, 1997).World culture defines how actors perceive problems, their own capacities and purposes; 'it sets parameters around what is proper or even thinkable in a given historical moment' (Cole, 2017: 91), which affects the nature of the adopted public policies.In other words, modern states do not choose from an infinite range of public policy options, but rather the range of available options is constrained by 'appropriate' models which fit with the current world culture and internationally promoted worldwide policy models.The detailed content of worldwide models changes over time (Meyer et al., 1997: 162).An important feature of the modern world culture is construction of the modern actor 'as an authorized agent for various interests (including those of the self)' (Meyer & Jepperson, 2000: 101).Individuals gain legitimate 'capacity and responsibility to act' for their individual or collective rights and benefits.This means that modern actors can mobilize more easily, also on behalf of others, including 'imagined potential actors' (Meyer & Jepperson, 2000: 108), such as unrecognized vulnerable groups and children.The internalization of world culture means expansion of social agency, legitimation of expanded actorhood and increased capacity for collective action.
World cultural models spread through various channels, to a great extent through international institutions.International institutions promote the global diffusion of these worldwide models of appropriate behaviour, encouraging national actors to adopt and internalize the underlying identities and agendas.According to world society theory, most states will tend to adopt the prescribed models at the high level of official statements and formal commitments, for example sign human rights treaties (Hafner-Burton et al., 2008).Often, this formal adoption of internationally legitimated models can be 'inconsistent with local <. ..> requirements, and cost structures' (Meyer et al., 1997: 154), leading to 'decoupling' between words and deeds.This de-coupling can occur when states declare their allegiance to certain policies, as a lowcost method to strengthen their international legitimacy (Hafner-Burton et al., 2008).Nevertheless, even such formal, high-level policy adoption can be significant, because the adoption of world models legitimizes new individual and organizational actorhood, and creates preconditions for social mobilization (Meyer et al., 1997), paving the way for deeper social change.
One advantage of world-culture explanations of social policy decisions is that differences in how social policy decisions are made in different political systems (whether they are results of collective political action or top-down directives) should not have much significance (Forrat, 2013).Fundamental ideas associated with the modern world culture (such as the centrality of the agentic individual) travel globally through multiple channels, and get internalized by different groups of the population -from service users and bureaucrats to dictators.Internalization of these ideas by different societal groups would provide a fruitful ground for the adoption of similar instrument choices, especially once related policy models are promoted.

Policy instrument choice
Policy instrument choice studies identify a wide range of factors that can affect the selection of instruments.Generally speaking, governments have a wide range of instruments at their disposal, and their selection can depend on their relative functionality and the cost of enforcement (Howlett & Ramesh, 2003: 200).It is, however, widely observed that the choice of particular instruments is also restricted by other factors, such as the organizational setting of policy makers and implementing agencies, the problem situation (Howlett & Ramesh, 2003;Linder & Peters, 1989) and macro-context, such as domestic institutions, the level of socioeconomic development, the overall governance mode, and preferred implementation style (Capano & Howlett, 2020;Howlett, 2009;Lenschow, Liefferink, & Veenman, 2005;Linder & Peters, 1989;Tosun & Treib, 2018).Policy instruments are embedded in national political culture (Pollitt & Bouckaert, 2011: 46) and governance pattern, 'such as the type of state-society relations, the level of state intervention' (Lenschow et al., 2005: 806).This can often lead to variation in countries' chosen instruments.Importantly, states generally 'tend to choose instruments from their predominant policy instrument mix and preferred implementation style', when policy changes are required due to international obligations (Bouwma, Liefferink, Van Apeldoorn, & Arts, 2016: 214).In some situations, as has been previously observed in Eastern Europe, new instruments can be introduced 'outside the existing implementation style', which has been associated with 'historical turns, domestic pressure and a shift to new modes of governance' (Bouwma et al., 2016: 214).
The cognitive dimension of instruments is increasingly recognised.Instruments should not be understood as merely functional devices, whose selection is technical.Instrument choice should be treated as an ideational and a political issue, a 'matter of faith and politics' (Hood, 1986in Howlett & Ramesh, 2003: 199), driven by policy-makers' subjective preferences (Linder & Peters, 1989).Policy instruments bear social and political meanings, values, and worldviews (LASCOUMES & LE GALES, 2007).In their instrument choice, policy makers not only seek effectiveness or stick to past decisions, but also engage in 'the construction of a shared sense, a common acceptance' (Capano & Lippi, 2017: 269).Decisionmakers have their subjective 'perceptions of the proper 'tool to do the job'' (Linder & Peters, 1989: 35), based on their ideas about social control and appropriate ways of exercising it.These ideas about the relationships between the state, private actors, individuals can act as a switch between types of policy instruments, e.g. from 'command and control' to less interventionist forms of regulation by persuasion (LASCOUMES & LE GALES, 2007: 13).These ideas shape the country's political culture, and therefore one could expect to find certain types of policy instruments in countries with certain political culture.This argument was made for example by Lenschow et al. (2005: 810), who linked 'basic properties of countries', such as national culture and institutions, to 'the likelihood of them taking up different types of policy innovations from abroad'.In particular, they argued that policy instruments are embedded in the national political culture, which can find reflection 'in the country's preference of legally enforceable rules versus more participatory approaches' (Lenschow et al., 2005: 806).Here, institutional categories focusing on the role of the state in governing society, power and political plurality are viewed as reflections of dominant ideas of equality, agency, and individualism, and therefor treated as 'forms of the political culture'.Therefore, arguably, countries with an authoritarian culture may 'be hesitant in the <. ..> adoption of communication-based instruments', for example (Lenschow et al., 2005: 807).
Despite the richness of instrument choice literature, crosssector and cross-national variations in instrument patterns are still poorly understood, and '''comparative policy instruments research" is quite undeveloped' (Capano & Howlett, 2020: 5).One of the issues (and the focus of this study) is the extent to which the adoption of instruments is affected by globalization.The impact of globalization on domestic state policy designs is contested.Even in conditions of international pressures, policy choices are not predetermined: various elements of policy can be 'imported', including general policy ideas, goals, specific instruments (Dolowitz & Marsh, 1996).Policy adoption can be constrained by country's political and institutional systems, and often, best practice from elsewhere cannot be transferred to the country because of domestic political opposition (Howlett, 2019).Recent scholarship generally agrees that domestic factors are no less important in conditions of globalization, arguing that 'the source of many of the changes in the patterns of instrument choice found in contemporary society lies in the domestic rather than the international arena' (Howlett & Ramesh, 2006: 175).Even within the European Union, Member States have been shown to 'refrain from introducing policy instruments requested or promoted by the EU' (Bouwma et al., 2016: 219), giving little ground to expect that countries without such close ties would opt for the same instruments.
How do the findings of these studies apply to the adoption of childcare deinstitutionalization policy?To answer this question, let us first consider the international context of DI policy adoption, in particular how this policy has been promoted by international actors.

DI policy promotion by international actors
Childcare deinstitutionalization policy has been promoted by international actors at different levels.Since the nearly universal ratification of the UN CRC, the UN Committee for the Rights of the Child repeatedly highlighted the issue of child institutionalization through the regular state party reporting process, whereby the Committee examines the reports of state parties to the Convention and formulates its observations and recommendations.To support the implementation of the UN Convention on the Rights of the Child, the United Nations issued the UN Guidelines for the Alternative Care of Children (Assembly, 2010), which the UN Committee has also been actively referring to in its communication with state parties.
To understand what elements of DI policy have been promoted, one can turn to the UN Guidelines for Alternative Care, UNICEF documents, and the implementation handbook (Cantwell et al., 2012), which was written to help governments implement the UN Guidelines, 'outlining the kind of policy responses required, and describing 'promising' examples of efforts already made to apply them in diverse communities, countries, regions and cultures' (Cantwell et al., 2012: 3).
International actors have promoted the idea that the best setting for a child to grow up is a family, including for children with special needs, whereas large residential institutions with no attention to individual child's needs are inappropriate.While this may seem a commonplace today, the Soviet system was based on 'the persistent and deeply-rooted belief ''that institutional care was an acceptable À even an ideal À form of childcare"' (Carter, 2005in Petrowski, Cappa, & Gross, 2017).
Children with various health and mental disabilities 'were considered a social 'problem' that families were deemed incapable of dealing with', and 'the responsibility for such children was given to the State instead of trying to understand the causes of a child's difficulties or reasons for vulnerability and working with families in crisis to provide support services and assistance <. ..>These children were placed in large-scale institutions ' Petrowski et al. (2017: 4).This contrasts with the family-centred policies promoted by international organizations today.The abovementioned international documents use the language that places the focus on the 'family' nature of appropriate care options: acceptable care should be 'family-based' or 'family-like', and resemble a natural family as much as possible.It is recognized that in some situations residential institutional care may be unavoidable, but resort to institutional care should be made only in extreme, exceptional circumstances (Burke, 1995: 2;Cantwell et al., 2012: 44; General Assembly resolution 64/142, article 14, p. 4, article 21).Based on the idea that the family, rather than impersonal state-run institutions, constitutes the best environment for the child, they promote the policy goal of reducing child institutionalization in the region and transition to communitybased care.
In addition, more concrete policy instruments have been promoted.For example, the handbook contains direct recommendations, such as 'National policy should', and elaborates in detail appropriate methods of DI policy implementation, at a relatively low level, e.g.: 'Provide support including financial support so that foster carers and carers in family-based care appropriately care for children with disabilities.Where appropriate, children with disabilities should continue to receive support as they move into adult life'; 'Collect and analyse data and undertake research to better understand the needs of children with disabilities and other special needs and to inform alternative care services' (Cantwell et al., 2012: 37-38).DI policy has been advocated for not as an abstract idea or distant policy goal, to which countries need to find their own paths, but rather as a detailed 'package', which contains a toolkit for governments.The above documents show that all policy instruments, for which the data in this study were collected (see Sections 3 and 4), have been promoted and codified at the international level, including: -foster care, whereby children are placed in a 'professional' family, i.e. a family other than their own, which has been selected, trained and supervised; -community services, such as child day care centres, rehabilitation centres, and other similar forms of family & child support, which help reduce the perceived need for formal alternative care; -returning children to biological families; -developing non-institutional residential care options, i.e. small group alternative care options that do not have the 'institutional culture', which can involve downsizing and reorganizing existing institutions to make them more open to the outer world, more family-like and attentive to individual psychological and emotional needs of the child; -supporting independent living of graduates of childcare institutions; -case management as a gatekeeping tool (Cantwell et al., 2012: 70;Sammon & Burchell, 2018: 22;UNICEF and World Bank, 2003: 82).
Given this international context and the findings of above mentioned studies, how much DI policy convergence should be expected among analysed countries?Applying the logic of world society theory to the case of DI policy, we expect to observe policy convergence across countries -at the high level of issue conceptualization (provision of family environment perceived as a policy issue) and policy objectives (reduction of child institutionalization to only exceptional cases and transition to community care), as well as at the lower level of policy instruments, as long as these are promoted by interactional actors.At the same time, we expect this convergence not to go very deeply, but in many cases occur through symbolic commitments, not associated with any structural changes.Given the promotion of deinstitutionalization by international actors, the DI policy is a particularly good case to test the workings of the world society.
Applying the findings of instrument choice literature, one can expect that even in cases where countries have committed to DI policy at the high level of policy 'ends', governments will not necessarily choose the same policy 'means'.Their choice for instruments is likely to be path-dependent and restricted by domestic political traditions, which, nearly 30 years after the fall of the Soviet Union, now vary greatly across the ex-Soviet region.Hypothetically, governments could introduce a variety of measures to reduce child institutionalization: tax or criminalize child abandonment, close institutions and tax/oblige members of the extended family to take care of the child; introduce a moratorium on placing any children in institutions, thus leaving parents to cope with the problems; encourage local churches to provide alternative care (and thus relieve national budgets), etc.One can hypothesize that governments will choose those, more or less coercive, options which better suit the national political culture.At the same time, the research suggests that sometimes domestic restrictions may be lifted, given the 'cognitive' aspect of instrument choice.In other words, given that national decisions are shaped by subjective perceptions of what constitutes an 'appropriate' measure and 'the right thing to do', there is a window of opportunity for international epistemic communities to influence national instrument choices.Let us now turn to the data on DI policy adoption among 15 politically different states, to explore the convergence of their DI policies, in particular a set of internationally promoted DI policy 'ends' and 'means'.

Data collection
I have collected data on childcare deinstitutionalization policies in 15 ex-Soviet republics in Eastern Europe and Central Asia: Armenia, Azerbaijan, Belarus, Estonia, Georgia, Kazakhstan, Kyrgyz Republic, Latvia, Lithuania, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, Uzbekistan.The data were collected from two sets of sources: (a) UN CRC reports submitted by state parties between 1990 and July 2020, (b) official national documents, located through snowball search starting with documents in the Better Care Network Library of Documents (https://bettercarenetwork.org/library/library-of-documents).Most data were collected from UN CRC reports, which I read for all analysed countries.State parties regularly report to the Committee on the Rights of the Child, stating the measures they have taken to realize their commitments on the children's rights, which often include deinstitutionalization (Ulybina, 2020).In some cases, where the relevant information could not be found in these reports, I searched for documents about alternative childcare in the relevant country in the online library.Having found the relevant information in those documents (which could be for example NGO or UNICEF reports), I used the references in the text to locate the original official national document and made every effort to locate those national documents through national online databases, in order to verify the information provided in the report.Relevant documents (national programmes, strategies etc. in the field of alternative childcare) were collected from national government websites, using website search function.For some countries, such as Turkmenistan, public information is very scarce, so I relied on third-party evidence, e.g.UNICEF reports.It is possible that some information was missed, therefore where I could not find evidence that a country adopted a particular instrument, I labelled such cases as 'Not located' rather than 'Not adopted'.Lack of evidence should not be treated as evidence of the lack of a policy/instrument.
Two broad sets of data were collected, which refer to: (1) DI policy ends (perception of the problem and goals), and (2) DI policy means (instruments).To trace the perception of the problem, I looked for statements which used the following phrases (or close equivalents): (a) family-like care, family-like conditions, familytype, family-level, family-based, as similar to a family as possible; (b) institutions as a measure of last resort, institutionalization as an exceptional measure; (c) explicit use of the term 'deinstitutionalization'.Additionally, I looked for evidence of official programmes, strategies, or action plans specifically focussed on childcare deinstitutionalization.For DI policy goals, I looked for statements about reducing child institutionalization, or transition from institutional to community-base care.For DI policy instruments, I looked for statements indicating that the government is supporting, developing or intending to use the following instruments, in order to achieve its DI objectives: foster family care; family and child community services; reintegration of children in their biological families; developing small group residential care, downsizing and reorganizing existing institutions; introducing case management as a gatekeeping tool; and supporting independent living after graduation from institutions.

Cross-national convergence of childcare deinstitutionalization policies
As predicted by world society theory, all analyzed countries, arguably driven by international normative pressure on governments to adopt the deinstitutionalization principle, have joined the global DI policy trend (Ulybina, 2020).They adopted the DI perspective on the problems of children deprived of parental care, and made high-level commitments to deinstitutionalize their childcare (Table 1 below).As follows from Table 1, countries use very similar DI language, including the explicit reference to deinstitutionalization, which indicates the adoption of the concept and the overall new framing of problems relating to children without parental care.The adoption of specific DI-focused programmes and plans suggests that various issues, which were previously regarded as stand-alone issues (health, education, social assistance, rehabilitation, child protection), are now linked together under the DI umbrella.
Surprisingly, countries also adopted very similar sets of instruments (Table 2).According to the collected data, the DI instrument choice does not depend on the national political regime: authoritarian states adopt the same instruments as their nonauthoritarian neighbours.The national political culture does not appear to be an important factor for (non-)adoption of these internationally promoted DI instruments.
Source: data on policy instrument adoption -compiled by author; classification of countries by regime type -based on Democracy Index 2019 by The Economist Intelligence Unit (2020).
Another important finding is related to the nature of these instruments.First, they place the individual, with their agentic capabilities, at the centre of service provision and decisionmaking.Community-based family support services include child day care centres, rehabilitation centres, counselling services, outreach services to families at risk, psychosocial support etc, to help meet the specific needs of individual children.Promotion of foster family care and child reintegration into their biological family is often accompanied with financial and other support to the family.Graduates of childcare institutions are supported to live independently, through professional training and job placement, provision of housing and medical aid, legal and psychological consultancy.Case management is a highly important innovation in the region.During the Soviet period, the profession of a social worker did not exist (e.g., Lotko, 2018: 177).Now, these states, including the isolated Turkmenistan, have introduced social work, often in coop-eration with the United Nations, UNICEF and NGOs (Bilson, 2010).Case management approach is meant to divert children from institutions, and ensure that children have access to the relevant social protection mechanisms and social services.Social workers assess the individual circumstances of each child in a difficult situation, prepare individual plans of care and support, monitor service provision, etc.In doing so, the social worker 'must have the capacity to provide assistance in a sensitive and supportive manner to particular client populations based on knowledge of human behavior and well-developed observational and communication skills' (Bilson, 2010: iii).So the state, impersonated in a social worker, 'establishes helping relationships' and 'helps clients to function effectively' (Bilson, 2010: iii).In other words, by supporting child reintegration in their family, the governments implicitly recognise the individual's capability to provide better care than the state.By introducing case management, states shift the focus of service provision on the individual, with their needs and interests (moreover -an individual which used to be regarded as defect and a burden for a family).By developing foster family, the state expects individual actors to become agents -agents for their own interests as foster parents and for interests of vulnerable children.In doing so, authoritarian states demonstrate their belonging to the modern world culture (Meyer & Jepperson, 2000) and, at the same time, authorize new types of agency in their countries.Parents, foster parents, children from socially disadvantaged families, children with special needs, social workers are legitimated as agents for their individual and collective rights, as individuals with 'capacity and responsibility to act' (Meyer & Jepperson, 2000: 102) and to make decisions.Individuals are put in a position of agents, who need the state as an enabler, a source of support, rather than guardian.
Second, these instruments are designed to support and gently direct, rather than prohibit or impose sanctions.They do not rely on the 'use of coercive authority to <. ..> force society's members to abide by government intentions' (Howlett, 2009: 81), as would be the case for example with criminalization of child abandonment or imposing legal responsibility for the child on extended family.Instead, these instruments show a preference for 'softer' approaches, when the state relies on financial incentives (e.g.supporting foster parents), communication (e.g.reintegration and gatekeeping), changing the nature of state services (e.g.retraining staff into case management approach), and reforming existing institutions (e.g.transforming institutions and supporting smaller care homes).More punitive instruments are also used, such as recovery of maintenance of children without parental care from their parents.However, UN CRC reports do not mention any increases in these charges or any efforts to improve enforcement of these measures.Countries do not seem to strengthen such coercive measures, but rather focus on non-coercive, liberal and individual-centred methods, as advocated by UNICEF.By adding these new instruments to the policy mix, governments set precedents for future non-coercive policy decisions.These findings are important not only for globalization and instrument choice research, but also for the growing scholarship concerned with political dimensions of social policies and social rights.Comparative research on social policy has so far produced mixed results about whether political regime matters for social welfare (e.g., Haggard & Kaufman, 2008).Many studies have focussed on the 'dark side' of authoritarian social policies, showing how authoritarian regimes used welfare policies and access to social rights as tools of social control, in order to ensure the regime's domestic legitimacy and control political behaviour of the population (Donno & Kreft, 2019;Forrat, 2013;Kaleja, 2017;Nova, 2019;Stubbs & Lendvai-Bainton, 2020).Social policies have been shown to provide tools for political coalition-building, a resource that authoritarian regimes can capitalize on.This study shows an unobvious development of authoritarian policies, in that it shows a spread of non-coercive policy instruments, which carry potential to empower individuals and authorize their individual agency.The findings indicate that globalization forces can prompt governments to adopt policy objectives and instruments, which do not serve as tools of political pressure or cajoling the population.Unlike for example recent demographic policies in Hungary (Nova, 2019), the analysed policy instruments are liberal in that they are not designed to generate more dependence on the state.They do not tie access to services (e.g., day care centres, case management, etc) to some politically desired behaviour.Moreover, they are 'modern' and 'Western' in the sense that they are 'enabling', underpinned by a rights-based, rather than charity-based, approach, which strives to put an individual with their particular capabilities, needs and preferences at the centre of social support.
Countries in the region are at different (and often very early) stages in their DI transition (Jones, 2019).It remains to be seen how fast the countries will progress in implementing the chosen instruments, and what impact their political systems will have on the nature of their implementation process.The adop-tion of such instruments as case management indicates that high-level policy commitments made by authoritarian states do not have to be 'empty promises' (Hafner-Burton et al., 2008).The existence of programmes to develop case management and overhaul the system of childcare provision shows that governments have moved far beyond formal, symbolic policy commitments.
These developments also pose the question whether current instrument choices can have an effect on the consequent policy and political development.It is recognised that policy instruments reveal and shape the nature of 'the relationship between the governing and the governed' (LASCOUMES & LE GALES, 2007: 3), which can eventually lead to political effects, beyond the original objectives they pursued.
The nature of the collected data does not allow us to elaborate on the nature of the mechanisms leading to policy convergence.One could hypothesize that governments chose these particular instruments rationally, as a result of cost-benefit calculations, expecting that the new arrangements will save budgetary resources.However, it is well known that the transition to these new institutions (community-based care, setting-up social work services, including case management, etc) is likely to be costly (e.g., Csaky, 2009: 12).
The fact that the same policy elements can be found in international recommendations and national policies is no indication of cause and effect.The cited international documents are rather late (2010 for UN Guidelines and 2012 for the implementation handbook).The implementation handbook is a collection of best practices, aggregating the international experience of the previous few decades in the field.So it is likely that early policy adopters chose some of these policy instruments independently, and their implementation was deemed successful and promoted among later adopters through international channels.On the other hand, the discussed policy objectives and instruments (downsizing institutions; reduction of child institutionalization through case management, provision of foster care and other community-based services) had been promoted by UNICEF long prior to the adoption of these documents (e.g., Burke, 1995;UNICEF and World Bank, 2003).Given these normative pressures, it is likely that the observed convergence is not coincidental, pointing towards globalization forces and neo-institutionalist cultural explanations of policy change.Some explanations of the observed convergence can be found in the scholarship investigating social governance patterns in authoritarian regimes, and in particular looking at the trajectories of social welfare transition in the post-Soviet republics.These studies highlight the 'nonprofitization' of the welfare state in the region (Salamon, 2015) -an increased agency of NGOs as advocates and service providers, a variety of new platforms and financing mechanisms for interaction of authorities with non-state actors.The increased willingness of authorities to cooperate with nonprofits in the social sector is understood to be a response to the growing burden of social problems, the relatively poor capacity, performance and legitimacy of the state in dealing with these issues (Aasland, Kropp, & Meylakhs, 2020;Cook, 2015), and an opportunity to share responsibility in an increasingly challenging policy environment among a broader set of actors (Holm-Hansen, 2018).Although child welfare policy in the region often remains largely in the realm of top-down government, public and private actors do engage in horizontal, network governance-like settings, with opportunities for non-state actors to set the agenda and convey ideas from outside the system (Bindman, Kulmala, & Bogdanova, 2019: 5;Holm-Hansen, 2018).Many studies indeed note the 'international inspiration' of social policy reforms in the region (Holm-Hansen, 2018: 144), strong international linkages of some civil society actors to intergovernmental organizations and transnational NGO networks (Froehlich, 2009;Pape, 2014: 205).These observations of the important roles of NGOs, including those with transnational ties, are consistent with world cultural explanation of policy development, and our findings suggesting certain penetration of liberal and individual-centred world culture in authoritarian regimes.

Conclusion
The results of this study indicate that authoritarian states in the ex-Soviet region do not differ from non-authoritarian governments in their choice of modern, non-coercive instruments for childcare deinstitutionalization.The data show significant convergence among 15 ex-Soviet republics in the adoption of DI policy 'ends' and 'means', i.e. countries see issues of children deprived of parental care from a DI perspective, proclaim similar DI policy objectives, for which they use similar DI policy instruments, despite drastic differences in political regimes.
The findings support the predictions of world society theory, which explains widespread assimilation of national policies through the pervasive influence of world culture and a key role of international organizations in promoting 'modern' policy models.By adopting the analysed DI instruments, governments signal that they expect their citizens to be pro-active, take initiative and volunteer as foster parents; they expect social workers to make their own judgement about the most appropriate solution for each child; they expect biological parents rather than the state to provide better care. . . in other words, they recognize the agentic individual.Such instruments drive forward a modern representation of the 'orphans' problem as a problem of securing children's rights, providing individualised support, equal opportunity, and enabling environment, rather than a problem of medical treatment and welfare.
Relevantly to the scholarship on authoritarian social policies, this suggests that world culture can penetrate societies quite deeply, even those with authoritarian regimes and apparently closed to external influences.Importantly, authoritarian governments do not proclaim a DI policy as an 'empty promise' but go further to introducing the internationally-approved instruments, which involve significant reform of social care provision, thereby autho-rizing new types of individual agency and providing new space for actors to defend their social rights.
For the international development community, the findings point at the complex, multi-aspect nature of inter-country convergence, as well as non-obvious relationships between country's domestic characteristics and cross-country policy convergence.We know that convergence in quality of life (health, rights, etc) can occur at the same time as incomes diverge (Kenny, 2005).The case of childcare deinstitutionalization shows that convergence in social rights does not have to be tied to convergence of political regimes.Furthermore, the data suggest that DI policy has been promoted by international actors as a 'package' including DI policy 'ends' and 'means', and DI policy also appears to have been adopted as this 'package' by analysed countries.This indicates that promoting policies as complete solutions can be effective.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Table 1
Adoption of deinstitutionalization 'ends': perception of the problem and adoption of DI policy goals by 15 ex-Soviet countries, as of July 2020.

Table 2
Adoption of DI policy instruments by country, by regime type.