Migrant Agency in an Institutional Context: The Akmola–Astana Migration System

Abstract This article addresses one of the key challenges facing transitional and emerging economies: managing rural–urban migration to tackle rural decline and the associated rapid urbanisation. We introduce New Institutionalism as a novel conceptual framework to analyse the interactions between the institutional environment and migrant agency in a rural–urban system: the Akmola–Astana migration system in northern Kazakhstan. Our results suggest that the government might be more successful if it engages migrant agency and incentivises remaining in rural areas instead of designing policies to discourage rural–urban migration.

and drawing on empirical data from the Akmola-Astana migration system in northern Kazakhstan. 1ur focus on rural-urban migration is timely, given the rapid urbanisation that is currently taking place in many transitional and emerging economies, especially in post-Soviet states. 2 While urbanisation is frequently credited with accelerating economic development and societal well-being, underlying rural-urban migration processes also entail societal costs.On the one hand, rural out-migration has been linked to the loss of young human capital, known as 'brain drain', and an associated demographic shift to the ageing of the rural population.On the other hand, urban migration can put a strain on affordable housing and the provision of public goods (Massey et al. 1998, p. 48).This has prompted some governments to design regulatory policies to manage rural-urban migration (Beauchemin & Schoumaker 2005, p. 1129).These policies usually focus on either enabling or constraining migrant agency by providing incentives for potential migrants to remain in rural areas or by deterring them from moving to cities (de Haas 2011, p. 6).Since its independence in 1991, Kazakhstan's government, for instance, has been experimenting with various policy interventions to regulate internal migration flows, initially encouraging migration into Astana, its new capital, founded in 1997, and later trying to halt it.Astana experienced a period of rapid urbanisation, more than tripling its population from around 300,000 inhabitants in the 1980s and most of the 1990s to one million in 2018.The highest share of incoming migrants to Astana originated from the surrounding Akmola province (StatKaz 2019).
It is now well accepted that structure/institutions and agency both shape migration processes and that neither has superior explanatory power vis-à-vis these processes. 3In fact, focusing on the individual agency to migrate without considering the institutional environment that aims to shape migration, and vice versa, runs the risk of missing the complexity of migratory processes (Castles & Miller 2013, pp. 28, 30).Simultaneously considering the interrelated forces of institutions and agency provides for a much more nuanced understanding of how migration processes might be influenced and shaped (Lacroix 2014, p. 671).Still, studies on rural-urban migration do not often explicitly consider institutions and agency simultaneously in their analyses and the interaction between them (Bakewell 2014, pp. 306, 309).Yet, there is a need to understand the complex interactions between individual (migrant) agency and the political, economic and social environment that facilitate or constrain agency in order to understand how policy measures affect migration processes.To fill this gap, we apply New Institutionalism to better understand migration processes, as it provides an opportunity to examine migrants as social agents who manoeuvre their way through complex institutional incentives and constraints while exerting, at the same time, pressure on institutions to change.While New Institutionalism has been widely used, it has not, however, been applied to the analysis of migration processes.Although there are migration research frameworks that incorporate institution and agency interaction, such as translocality (Greiner & Sakdapolrak 2013, p. 375) or migration systems (Bakewell 2014, p. 306), New Institutionalism is, in our case, an ideal conceptual framework as the core of the theory focuses on a changing institutional environment.Valid reasoning about people's agency in a migration system should include the investigation of both ends-the sending and receiving locations of the migration system and the linkages between them (Castles & Miller 2013, p. 27).This is even more true given that the livelihoods of most migrants and migrant households are translocal (Thieme 2008b, p. 67). 4 Our central research question is, therefore, how does the institutional environment affect people's agency in the sending and receiving areas of the Akmola-Astana migration system and vice versa?
In the next sections, we first outline the merits of New Institutionalism as a conceptual framework and demonstrate how it helps to understand the interaction between institutions and agency in migration processes.Second, we describe the evolution and transformation of Kazakhstani migration institutions, unpack individual migration decision-making processes and look at how the agency of (potential) migrants has been framed by existing state interventions.We will show, for instance, that in reaction to the Kazakhstani government artificially raising urban housing costs through interventions in the Astana housing market, migrants responded through collective agency by taking advantage of their family networks.Urban relatives provided accommodation and rural relatives financially supported their migrant family with so-called reverse remittances.This and other results suggest that instead of constraining migrants' agency, promoting institutions that expand the agency of the rural population to stay is more effective in moderating rural-urban migration.

The merits of new institutionalism in framing migration processes
Although we can look back at decades of research on international and internal migration, no comprehensive theory is available to explain migration processes such as deciding to migrate or not.A number of renowned scholars, such as Portes (1997, p. 811) and Castles (2010Castles ( , p. 1582)), even argue against the idea of an all-embracing theory for migration studies.They suggest that migration research should use middle-range theories that can integrate the insights of various social sciences in order to improve the understanding of migration.De Haas (2010b, p. 241) acknowledges that these appeals correspond with a general paradigm shift in contemporary social theory away from grand theories and towards hybrid approaches that can integrate a range of disciplines, paradigms and theories that are both flexible and disciplinarily neutral (King & Skeldon 2010, p. 1634).
A conceptual framework based on New Institutionalism makes it possible to merge general sociological and economic assumptions about institutions and agency with actual policies.New Institutionalism is at the junction of political science, sociology, history and economics and, as pointed out by March and Olsen, its 'spirit is to supplement rather than to reject alternative approaches' (March & Olsen 2006, p. 16).As such, New Institutionalism has great power to provide an integrative framework for complex research designs (Goodin & Klingemann 1996, p. 25).
New Institutionalism is able to theoretically embrace institutional genesis, reproduction and change, and it links individual agency at different societal levels (Thelen 2004, p. 31).At the macro-level, New Institutionalism can explain global and national processes of social, economic, political and cultural change while linking them to agency within an institutional environment.The institutional meso-level is attached to networks, communities and localities that are relevant for social interaction and link the macro-and micro-levels with each other (de Haas 2010a, p. 1591).
There are three well-established schools of New Institutionalism: historical institutionalism, sociological institutionalism and rational choice institutionalism (Olsson 2016, pp. 1, 22).Although the schools differ somewhat in their understanding of the mechanisms of institutional change, they are united in their theoretical core (Peters 2012, p. 184).Thus, in line with scholars as diverse as Jakimow (2013, p. 494), Lowndes and Roberts (2013, pp. 40-1), Koning (2016, p. 639), and March and Olsen (2006, p. 16), we combine elements from all three schools into one conceptual framework.
Figure 1 depicts our conceptual framework of New Institutionalism in the context of a migration system.Comprehensive consideration of all three schools extends our understanding of the ways in which institutions at different societal levels affect individual agency; in our case, the study of rural-urban migration decisions (Campbell 2004, p. xiv;Jakimow 2013, p. 499) but also path dependencies in migration systems, or positive and negative feedback loops, to better understand institutional change or stasis across time.Sociological institutionalism delivers an understanding of how a migration system interacts with societal norms, beliefs and ideas; how ordinary or elite actors within a particular institution foster or inhibit institutional change, on the one hand, and how institutions influence agency, on the other.Rational choice institutionalism allows us to understand the reasoning behind decisions related to migration processes; for example, subjective cost-benefit considerations, the effects of possession and the use of power.
As mentioned before, we follow an institutionalist view of agency where the institutional environment (both in sending and receiving areas) frames the ability of actors to make choices (Lowndes & Roberts 2013, pp. 16, 52).Agency possesses an iterative (based on past patterns of thought or action), a projective (based on perceived possible future trajectories of action) and an evaluative element (based on the practical and normative assessment of alternative actions) (Emirbayer & Mische 1998, p. 971).This implies that exercising agency involves evaluating a given situation, prior experiences and possible solutions, responding to uncertainties or challenges, inventing new possibilities, and mediating between and contextualising possible consequences.This institutionalist view of agency is classified by Lowndes and Roberts (2013, p. 106).We will use this systematisation of agency in our analysis of migrant decision-making in the Akmola-Astana migration system and use four different classifications of agency: cumulative agency, when the actions of many independent actors have an effect on an institution; collective agency, when actors work together under the same institutional environment; combative agency, when actors oppose other actors and their institutions; and constrained agency, when actors are always constrained to some extent.Although any one of these agency types may dominate in certain situations, they often exist simultaneously, demonstrating the diversity and overlapping agency of people (Coe & Jordhus-Lier 2011, p. 217).

Methodology
Our research uses an approach that mixes qualitative and quantitative data, collected between 2016 and 2017 in the Akmola province of Kazakhstan.This article draws on the 68 qualitative semi-structured interviews with potential and actual migrants, and policymakers.The interviews were conducted by a lead researcher, who was assisted by domestic research assistants.Interviews were held either in Russian or Kazakh and took usually 30-40 minutes. 5We conducted interviews with 23 government officials at different administrative levels and four migration experts and political scientists, who were purposely selected.Moreover, we interviewed 27 potential rural migrants in several villages in the Akmola province and 14 rural-urban migrants in Astana in order to understand the regulatory/policy environment as well as the agency of (potential) migrants. 6Interview participants were identified using a mix of random route sampling7 and snowball sampling.The random route sampling was used to seek out people with different life experiences, perspectives and characteristics.In remote villages, interview participants were also asked for their assistance in finding additional interview partners, that is, individuals with migration experiences or with a migrant family member.A list of all interviews referenced in this article can be found in Table A3 in the Appendix.
This article also draws on data from a quantitative survey of 400 rural households (potential migrants).The qualitative research further served as the basis for the design of the household survey.The quantitative survey of rural households followed a three-stage clustered random sampling procedure (districts, villages, households).In each village, ten households were randomly identified via random route sampling, as household lists were not made available.Within the households, the person between the ages of 16 and 50 who had most recently celebrated a birthday was interviewed.On average, interviews took 80 minutes and were conducted in Russian or Kazakh depending on the preference of the respondent.In these interviews, relevant data on all adults, and on the general socio-economic situation of the household, were collected.Survey questions aimed at understanding the institution-agency interaction of rural dwellers who intend to stay put or to migrate.

The Akmola-Astana migration system
During Soviet times, notable parts of the population moved to and from Kazakhstan.After the collapse of the Soviet Union, many migrants, most notably ethnic Germans and Russians, returned to their place of origin or titular states.In 1990, the population of Kazakhstan was estimated to be 16.3 million.Emigration related to the dissolution fall of the Soviet Union caused this number to drop to 14.9 million in 2003.To counter the losses, in the late 1990s the Kazakhstani government initiated a return programme, the Oralman Programme, to encourage the return of ethnic Kazakhs living, for example, in Mongolia, Uzbekistan or China.In the years that followed, the Kazakhstani economy became stronger, leading to significantly less out-migration and a slightly higher return of ethnic Kazakhs to their home state.This resulted in a positive migration balance that has been ongoing since 2004.Combined with a high birth rate amongst ethnic Kazakhs, the total population grew to 16.3 million in 201016.3 million in and to 18.3 million in 201816.3 million in (StatKaz 2019)).

Institutional environment of the Akmola-Astana migration system
The institutional environment of the Akmola-Astana migration system can be separated into two categories: urban planning policies, especially in the context of nation-building that revolved around the development of the new capital Astana (see Table 1); and rural and regional development policy measures such as building infrastructure and regional education and health facilities aimed at slowing the rural exodus (see Table 2).Internal passport and registration policy (Historical institutionalism) The former Soviet Union used an internal passport system and city registration to regulate population movement and urbanisation (Osmonova 2016, p. 237).After Kazakhstan's independence, this institution was reformed to allow registration by current address.Registration at the place of residence is still mandatory but can no longer be denied.If a person owns a home or shows a valid rental contract at the place where they wish to register, registration is granted.If not properly registered, new urban migrants are, however, often harassed by the police (Yessenova 2005, p. 670).*Nevertheless, most potential and actual migrants perceive this to be a nuisance rather than a deterrent to rural-urban migration.Therefore, many people in Astana remain unregistered despite forgoing potential benefits associated with registration, for example, public services such as free healthcare (Sanghera et al. 2012, pp. 15, 28).This in turn fosters the translocality of migrants as they have to return to their place of registration to access certain public services.

Astana
'city of modernity' and political/cultural centre (Sociological institutionalism) The Soviet narrative that cities are the cradle of modernisation and progress is still in effect and perpetuated in today's Kazakhstan (Alexander et al. 2007, p. 2).The narrative promoted by the government is that Astana is 'catching up with the world' and provides a chance for ordinary Kazakhs to participate in modernity (Anacker 2004, p. 531;Laszczkowski 2016b, p. 149).About half of the villagers in the household survey stated that compared to cities like Astana there was no social or cultural life in the village and that their children would have a better life in the city.Moreover, attractive new higher education facilities were established in Astana, which culminated in the establishment of the Nazarbayev University, probably the most prestigious university in Kazakhstan; see, for example, Koch (2014b, p. 51).Frequently, students who move to the city for education develop aspirations for an urban career and lifestyle.This weakens their familial and emotional attachment to their rural home region.Thus, many do not return (Buchenrieder et al. 2020).Therefore, the Astana migration system is characterised by the accelerated rural-urban migration of young people from the countryside and a lack of well-educated young professionals and an over-ageing of the sedentary population in rural areas.About three-quarters of the household survey respondents stated that, in cities like Astana, their children would have access to better education facilities and 40% of them stated that lack of education facilities is a major constraint to remaining in situ.†

Urban job market (Rational choice institutionalism)
Moving the capital to Astana created a vibrant job market with relatively high salaries compared to the surrounding rural areas.Potential migrants in rural areas often acknowledge in both qualitative interviews and in our household survey ease of finding high-paying jobs in Astana.About half of the household survey respondents stated that they believed their career prospects would improve and they could achieve a higher standard of living in the city.
(Continued ) Moving the capital city from Almaty to Astana lies at the heart of Kazakhstan's official nation-building project (Anacker 2004, p. 515;Bekus 2017, p. 806;Caron 2019, p. 183).The government transformed the narrative of a multi-ethnic country into one of a country for (ethnic) Kazakhs.This includes recognising Kazakh (together with Russian) as the national language (GovReKaz 1997;Caron 2019, p. 201).
(Sociological institutionalism) Kazakhification is an unofficial (the official discourse describes it as 'harmonisation') but on many levels observable policy, for example, through the renaming of streets referring to Kazakh traditions or historical figures and the dismantling of Soviet monuments (Bekus 2017, pp. 797, 800;Caron 2019, pp. 186-87, 196).More importantly, speaking Kazakh has become more relevant in businesses and, in particular, within the public administration (Wolfel 2002, p. 501;Peyrouse 2007, pp. 484-85;Bissenova 2017, p. 652).Public employees must be able to speak both languages at a sufficiently high level.However, only 10% of ethnic Russians surveyed stated that their level of Kazakh was sufficient to find a job in the city.Thus, there is a strong rural-urban migration constraint along ethnic lines.This is also reflected in the fact that ethnic Russians originating from the countryside and studying at higher education facilities in Astana have a much higher return intention than ethnic Kazakhs (Buchenrieder et al. 2020).

Housing market/ city development (Historical institutionalism)
Initially, migrants were welcomed to fill the new remote capital city, Astana.Following a slow start, Astana grew quickly until the early 2000s, when the city government found itself needing to restrict the influx because the urban infrastructure could not keep pace.As the City Planning Department regulates the designation of building land, construction planning and building permissions, it determines the size of the housing market.By rationing the number of construction permits for new housing, it intentionally drove housing prices up.About 70% of the household survey respondents stated that they could not find affordable housing in the city (see Table A1 in the Appendix).‡

(Sociological institutionalism)
As a counterbalance to the high property prices and rents, the government set up housing programmes for state employees.These subsidised housing programmes allow people with low-income occupations to gain access to affordable housing in Astana (Bissenova 2017, p. 644).However, the process can take up to seven years from application to allocation.¶ (Continued ) To some degree, a number of these policies are at odds with each other and risk creating a contradicting institutional environment for migrants.Astana continues to grow8 and is being promoted as a city designed for more than three million inhabitants, as declared by Nursultan Nazarbayev (President of Kazakhstan 1990-2019) in 2016. 9This may further spur ruralurban mobility.There is also an understanding amongst government officials and experts that it may be impossible to create equal living conditions across all regions of Kazakhstan and that some remote villages are beyond saving and further urbanisation should be promoted.10Nevertheless, senior politicians and government officials see unregulated urban growth as a major problem that will increase the burden on social infrastructure, especially healthcare and education.11Moreover, government elites may be even more concerned about dissatisfied urban masses living in precarious conditions close to the centre of power.Once a critical mass of angry youth is reached, this could have the potential for regime-changing protests.Hence, there is a strong desire to regulate urban migration. 12igrant agency in the Akmola-Astana migration system The government's portrayal of Astana as a shiny modern metropolis, combined with the fact that the capital city is an economic and political power centre, continues to motivate rural residents to migrate to Astana (cumulative agency).As mentioned above, this massive migration influx has taken a toll on urban infrastructure, which, from 2003 onwards, prompted the city government to create policies (starting after 2003) aimed at repelling in-migration; for example, by limiting free health care only to registered city citizens. 13One way to cope with the high housing costs is to In response to over-ageing and the lack of human capital in rural areas, special scholarship programmes were set up.These programmes offer scholarships for young rural adults to study professions in demand in rural areas.Graduates are attracted to rural areas, in a nationwide initiative, by the offer of higher salaries and subsidised housing.Participants of these programmes are required to work for five years in rural areas (GovReKaz 2018).Furthermore, decentralised education facilities have been established and adapted to regional demand.In Akmola this has led to the opening of colleges in regional towns, mostly in the following professional fields: agriculture, mechanical engineering and primary health care.Thus, young adults do not have to leave their home region for higher education, which may even lead them to stay in the region after graduation.The decentralisation of education facilities also creates local jobs and attracts new businesses (Buchenrieder et al. 2020  stay in precarious accommodation without proper registration.14Landlords became aware of a lucrative business opportunity and began renting out single apartments to unusually high numbers of migrants, sometimes up to 20 people.However, this practice was prohibited in 2016 through a government regulation stipulating a minimum of 15 square metres per tenant. 15Despite these state interventions, migrants continue to flow into the capital in the hope of finding better paid jobs or starting a business.Most of them have relatives or acquaintances already living in Astana who help them to find their first jobs.These jobs are usually below their level of education, often in the service or construction sector.In many interviews, including 16, 18, 19, 20 and 21, respondents noted their intention to bring their families to Astana but admitted that, at the time, they were only able to take care of themselves and were struggling with the precarious income-cost ratio.This job mismatch may also have been caused by a rather naïve belief in the supposedly endless job and business opportunities in Astana, as shown in the example of Rustam. 16Rustam was a student in his final year of mining technology at a regional college and planned to work in the gold mine in his home village.However, he wanted to move to Astana in about five years, because 'there are so many job opportunities'.It was not likely, however, that Rustam would find a job in Astana that corresponded with his academic qualification. Collective migrant agency-actors working together in the same institutional environment -is shaped through collective decision-making within families because individual migration may affect the whole the family.Almost all respondents in the rural household survey agreed that migration decisions were made jointly within the family.17One of our interviews provides some further insight here. 18Madina was an ethnic Kazakh from Mongolia who had moved to Akmola ten years ago.She worked as a system operator in a local gold mine and lived with her three daughters.Together they decided that, in the medium to long term, they wished to move to an urban centre and planned their professional lives accordingly.The eldest daughter worked with Madina in the mine; the second eldest attended a college in Astana and the youngest was still going to high school in the village but planned to study in Astana as well.The eldest daughter had already tried her luck once in Astana but failed: she was unable to find a job with a salary high enough to cover the exorbitant cost of living and, lacking a supportive family network in Astana, was forced to return to the countryside.However, the whole family intended to move together to a city, preferably Astana, as soon as the youngest daughter graduated from high school.This example highlights the translocal characteristic of migration processes and the safety net function of families at the place of origin. 19he joint decision-making of family members also relates to risk-pooling strategies, such as that of remittances sent by urban migrants to their families at the place of origin.Like in other parts of Central Asia this exemplifies the strong translocal links between different locations (Thieme 2008a, pp. 327, 338).However, unlike other parts of Central Asia, such as Kyrgyzstan, where transnational livelihoods are multilocal, spanning across countries and rural and urban locations (Thieme 2014 p. 140), in Kazakhstan they are mostly intranational covering rural and urban.Also, receiving remittances and return migration are far less common in Kazakhstan, and there is a reverse flow of remittances from rural to urban areas. 20Around one third of the respondents in the household survey sent reverse remittances compared to only one fifth who received them.Furthermore, about four times more money was transferred to urban migrants than was received by rural households.This is relevant as urban income levels often do not match migrants' expectations (Osmonova 2016, p. 239). 21According to Dietz et al. (2011, p. 23), a large share of migrants in Astana reported their income to be the same as or even less than before they migrated.It is more likely that family members who stay behind will later join the migrant, for example, after they retire, and not the other way round.In our qualitative interviews, villagers reported the intention to reunite with their adult children in urban areas once they were old or had retired. 22Overall, 25% of the rural household survey respondents stated that there was the possibility that they would move at a later point in their life.
The case of Sergei and Maria23 illustrates combative agency (opposing other actors and their institutions).Born in a small village in Akmola province, they were both ethnic Russians.Neither had received a higher education.Soon after Astana became the new capital, Sergei was drawn to the city and worked as a day labourer on construction sites before moving on to work for an air conditioner maintenance company.The owner started a side business producing advertising material such as small give-aways made from plastic or metal but also huge banners covering whole building facades where Sergei began to work.After a while he left the company and set up his own business producing advertising material.Because he did not speak Kazakh, the municipality and government departments, however, are no longer hiring his company.Public jobs require a certificate of proficiency in Kazakh (Peyrouse 2007, p. 485).This does not officially cover government contracts for external service suppliers.However, government officials lean towards applying this regulation informally in government tenders.The couple contemplated leaving for Russia but decided against it.At the time of interview, Sergei's company catered to the private sector and employed around 20 people.The couple and their two children counted as upper middle class in Astana.Despite losing contracts because of the government preference for Kazakh speakers, the company managed to adapt its customer base and the couple remained in Astana.Similarly, Blackburn (2019, p. 227) provides evidence that Russian speakers are pushed towards the private sector in Kazakhstan.
Constrained agency refers to the fact that actors are always faced with constraints that affect their agency.This is of course also true for (potential) migrants, who may find their intentions to move or stay constrained for various reasons.In our household survey, many respondents stated that the lack of good education facilities in rural areas was a major constraint to remaining in situ (see Table A1 in the Appendix).An example of this is the case of Nurlan, who worked as a security guard in Astana.Previously, he had lived and worked in Baikonur and his family had no intention of moving.However, after his wife visited Astana, she was convinced that the city would offer the best educational facilities and prospects for their children, and they moved to Astana. 24In contrast, a substantial number of the household survey respondents thought of themselves as 'forced stayers' (25%).Many of them stated that they would find it difficult to move because they had to take care of family members, including elderly relatives who were unwilling or unable to move.Even though nursing homes existed in both rural and urban areas, almost all of our household survey respondents said it was still socially unacceptable to place elderly relatives into these homes.
Nevertheless, the majority of our household survey respondents (68%) did not wish to move and were voluntary stayers.Most were quite content with their life in the countryside and believed that with hard work they could still make a decent living in the village (see Table A1 in the Appendix for details).Even if the migration system did not encourage them to stay, neither did it force them to move.Many people were also emotionally attached to the countryside.They preferred the rural lifestyle and appreciated the natural environment and the traditional social norms, as shown by statements such as, 'Nature is so beautiful'; 'I love fishing and hunting'; and 'All my friends and family live here'. 25Not surprisingly, 71% of our household survey respondents stated that they enjoyed the rural way of life and almost 90% mentioned that growing their own food was important to them. 26However, despite this positive perception of life in the countryside, many respondents wanted their children and grandchildren to move to the city because life there was perceived to be more comfortable and to have better educational and job prospects (see Table A1 in the Appendix). 27This was frequently the starting point for relocating the whole family.In our interviews, we repeatedly heard the statement: 'We 24 The name has been changed to maintain the anonymity of our interviewee (interview 30, Nurlan, Astana migrant, Astana, 2 March 2017).
27 Interviews: 3, village mayor, Stepnogorsk district, 27 May 2016;19, villager, migrant household, Shortandy district, 17 September 2016;26, migration expert, Astana, 28 May 2016;33, villager, migrant household, Shortandy district, 16 September 2016;34, villager, migrant household, Shortandy district, 16 September 2016;35, villager, migrant household, Shortandy district, 18 June 2016.will stay until our children finish school (secondary/high school), and then we will all move to the city'. 28rban areas located in north Kazakhstan were heavily Russified during Soviet times, with the result that many urban Kazakhs speak better Russian than Kazakh (Wolfel 2002, p. 490).Although Russian is still important as a business language, the political environment is such that fluent command of Kazakh is a demanded skill, and one can observe a change in language use in favour of Kazakh, especially amongst young Kazakhs (Smagulova 2016, p. 101).As mentioned before, public jobs require a certificate of proficiency in Kazakh and private companies in urban centres often seek native Kazakh speakers as contact persons for state authorities. 29This is shown in the case of Alexandr, an ethnic Russian.He studied accounting in Astana but could not find a job in his profession as he lacked the necessary Kazakh language skills.His friends from university all remained in Astana even if it meant having to take on jobs unrelated to their education, which he refused to do.For this reason and because he liked the rural lifestyle, he decided to return to his village. 30At the time of our fieldwork it had been roughly 20 years since the issuance of Law No.151-I on languages that defined Kazakh as the state language (GovReKaz 1997).Despite free language courses offered by the government, however, ethnic Russians overwhelmingly had not learnt the language.Nevertheless, many Russians consider Kazakh-language proficiency to be linked to better economic opportunities (Smagulova 2016, p. 101).It is therefore possible that lack of Kazakh proficiency deters some ethnic Russians from moving to the city.Studies suggest several reasons for this linguistic shortfall, including the negative feelings of many Russians in relation to a perceived loss of status as the ethnic majority and political elite (Laitin 1998, p. 155;Wolfel 2002, p. 501;Blackburn 2019, p. 224).Nevertheless, Russian-speaking urban Kazakhs still have a very positive attitude towards Russianspeaking minorities (Blackburn, 2019, p. 230).
This interpretation is supported by our finding that ethnic Russians were more inclined to stay in the countryside compared to ethnic Kazakhs.Only 15% of ethnic Russians in our household survey intended to move within the next three years compared to 30% of ethnic Kazakhs.Similarly, findings by Aldashev and Dietz (2014, p. 393) indicated that ethnic Russians were less mobile within Kazakhstan.This may in part be explained by the fact that ethnic Russians who emigrated from Kazakhstan after the collapse of the Soviet Union were mostly young people.Thus, the ethnic Russian population is on average older than the ethnic Kazakh population (Peyrouse 2007, p. 493) and it is well known that older people are less mobile (see Oh 2003).However, at the same time, ethnic Russian university graduates interviewed in a quantitative survey had a much greater intention to return to the countryside than their Kazakh counterparts (Buchenrieder et al. 2020).Also, fewer ethnic Russians in our household survey have supportive family networks in Astana (only 16% compared to 33% of ethnic Kazakhs) and these networks are also smaller (three relatives for ethnic Russians and ten relatives for ethnic Kazakhs).This may result in less family support for new arrivals and also affect the reverse remittances sent.Although ethnic Russians in our household survey send fewer reverse remittances (as fewer of them move), the average amount they send is substantially higher (about 20%) compared to the average amount sent by rural ethnic Kazakh families.If our results hold true also for a larger sample, this could mean that the Kazakhification of the northern countryside might reverse in the long run.However, the ethnic Russian population is still shrinking due to low birth rates and over-ageing (Laruelle 2018, p. 71), which may offset the fact that more ethnic Kazakhs from the north are moving to the cities.

Discussion of interaction results between institutions and agency in the Akmola-Astana migration system
Kazakhstan was the only Soviet successor state whose titular group was an ethnic minority (Schatz 2000, p. 489).The most important task of the newly independent government was thus to establish a Kazakh conception of nationhood in which the ethnic composition of the population played an important role.The goal was to increase the ethnic Kazakh population in Kazakhstan to above 50% (Kesici 2011, p. 32) and to claim the land as historically Kazakh territory.Today, ethnic Kazakhs represent the majority due to the emigration of ethnic Russians and other minorities, a high birth rate amongst ethnic Kazakhs, and the resettlement of returning ethnic Kazakh migrants (Alff 2012, p. 6).Regional imbalances do however still exist.Thus, the government promoted the internal relocation of ethnic Kazakhs into the northern regions, such as Akmola province, where traditionally Kazakhs had been a minority (Smailov 2011, p. 19).The foundation of Astana as the capital city in 1997 in Akmola province was a defining point for the Akmola-Astana migration system, one that significantly influenced internal migration processes in Kazakhstan.After Astana became the capital city, people moved there from all over Kazakhstan, above all, from the surrounding northern areas.Although, as pointed out by Anacker (2004, p. 524), it is an open secret that this was a geopolitical decision to finally claim the northern parts of the country as ethnic Kazakh territory, the formal justification for attracting Kazakhs to Akmola and, more precisely, to Astana was the creation of a focal point for economic growth in the northern territory of Kazakhstan (Anacker 2004, pp. 523, 528;Koch 2013, p. 144).A possible downside of nation-building and Kazakhification policies, although not in themselves anti-Soviet or anti-Russian,31 is that the agency of non-Kazakhs, especially ethnic Russians, could be constrained due to their often insufficient command of Kazakh.However, the Kazakhification process has slowed in recent years (Blackburn 2019, p. 217).Despite these constraints, ethnic Russians only complained about it in two of our qualitative interviews32 and, as pointed out by Spehr and Kassenova (2012, p. 146), the majority of non-Kazakhs are proud to be Kazakhstani (whether this holds true in contemporary Kazakhstan is uncertain; however, looking at the current invasion of Ukraine by Russia, we assume that this has rather strengthened any Kazakhstani emotion).
Transforming Astana into a capital city perpetuated older ideas.The government framed its narrative for Astana on the Soviet narrative of centralisation and cities as cradles of modernity (Laszczkowski 2011b(Laszczkowski , p. 96, 2016a, p. 61;, p. 61;Koch 2014a, p. 434).In general, Kazakhstani cultural life is centred on the urban world and rural life is rarely celebrated. 33Nevertheless, rural migrants and visitors are impressed by the spectacular architecture and the ongoing construction boom, a view shared by most Astana residents (Osmonova 2016, p. 241;Laszczkowski 2016b, pp. 153-54, 157).These institutional changes and the comparably high salaries strengthened cumulative agency by creating an enormous draw for people in the surrounding countryside, pulling them towards Astana.Nevertheless, there is often disappointment once rural newcomers are confronted with the realities of life in Astana, which often do not correspond with their ideals.Subsequently, they often underreport the harsh living conditions and high costs of living when talking with their rural family and friends. 34 Until the population grew to around 500,000 people in 2003, the city government of Astana welcomed the influx of migrants.After that point, the city government began to change certain public service institutions, for instance, only registered city citizens gained access to free health treatment or educational services because the development of the urban infrastructure could no longer keep pace with the influx of people.Officials from the city's migration department 35 referred to a situation of 'self-defence' because central government regulations were missing and therefore, they had to invent countermeasures. 36Despite calls by senior Kazakh politicians for more controls over internal migration, in the post-Soviet era, citizens have every right to choose their place of settlement, as stipulated in the constitution, and such a reversal to Soviet-style controls would be unenforceable. 37The City Planning Department was transformed from a service agency providing transport, education and health services into a gatekeeper institution (this informal and diffuse process is hard to pinpoint but it began at least since the year 2011).Since it controls development, land-use designation and planning permission, it effectively regulates the housing market. 38As a result, in spite of the construction boom, the housing market is rationed and Astana has, together with Almaty, the highest housing prices in the country (Osmonova 2016, p. 241).In this case, the high prices are a desired effect to reduce the influx of poorer internal migrants. 39As a consequence, the wage surplus paid in Astana compared to the surrounding countryside is often eaten up by the higher urban housing costs, given the shortage of affordable flats (OECD 2017, p. 55).To counterbalance the rising costs, the government set up a housing subsidisation programme for its employees mostly in the 2010s (Bissenova 2017, p. 644).There is, however, a long waiting list. 40Thus, residents in Astana often call upon the collective agency of their extended family to combat the severely rationed housing market.The rural-based family may offer financial support (that is, reverse remittances) and the urban-based family, if available, may provide accommodation, not only increasing the chances of migrants staying in the city but also reinforcing the extended family network itself.Poorer migrants without family networks are forced by these interventions in the housing market into precarious housing as Yessenova (2010, p. 21) has shown in her work on Almaty.
Another form of combative agency is to build illegally, for example, on family land, in the hope that these buildings will in the end receive a legal status.These illegal buildings, found frequently on the outskirts of Astana, are often of low quality. 41They are populated mainly by poorer people, who often face additional insecurity as these buildings, which were constructed without a permit, face the possibility of demolition orders by government officials.Another negative outcome is that corruption and fraud thrive in such an anarchical climate, as, for example, in the case of Aigerim. 42Her family sold all their property in the north and used the money to buy land close to Astana.They had several meetings with local government officials, who promised them a construction permit.After paying for the land, they applied for planning permission, as promised and learned that their area had never been designated for residential development.They now live in dire circumstances in a very small apartment elsewhere in the city.
upon by the longer-term residents while, on the other, their labour and the economic surplus they create are dearly needed. 43he collapse of the Soviet education system in the 1990s led to a dramatically reduced number of regional colleges (OECD 2007, p. 36).Only the large education facilities in the major cities remained intact (Toleubayev et al. 2010, p. 368), and were even supplemented by new facilities, offering programmes in crafts and trades (Buchenrieder et al. 2020).As a result, young people were drawn into bigger cities (cumulative agency) and often stayed on after graduation, as they grew accustomed to city amenities and their social bonds to their home region weakened.This migration behaviour is known as the 'migrating-to-learn'-'learning-to-migrate' chain (Rérat 2016, p. 279).Nevertheless, the increased rural-urban migration of young people results in a lack of professionals with higher education in rural areas, with flow-on effects for the quality of public service infrastructure in those areas; for example, in health and education.Chronic public underinvestment may also play a role.Thus, the elderly or families with older children in urban centres are also likely to relocate to towns as their agency to stay in the rural area of origin becomes constrained.
A reassessment of the education system took place in the early 2000s.In a reaction to counter the lack of well-educated professionals in rural areas, the central government set up scholarship programmes and decentralised higher education facilities (Buchenrieder et al. 2020).Moreover, the previous focus on urban development programmes was complemented by a more decentralised rural and regional development approach.However, the continuation of these rural and regional development programmes crucially depends on the national budget, which is heavily dependent on the price of fossil energy (Gallo 2021, p. 19).If the national budget is hit by an external shock such as low commodity prices-for example, for oil and gas exports-the policy pendulum may swing back to focus again on urban development, where the power elite resides.

Conclusion
It is important for migration studies to systematically describe the interaction of institutions and agency at the region of origin and destination when explaining migration processes.Our conceptual approach therefore embraces institutions and agency in an egalitarian manner by applying a New Institutionalism framework.Our case study is the Akmola-Astana migration system in Kazakhstan.This migration system is intriguing because the Kazakhstani government has proactively and reactively altered institutions in Akmola and/or Astana, thus altering the agency of potential and/or actual migrants.
Initially, the cumulative agency of rural people moving to Astana was welcomed and promoted by creating a narrative of Astana as a 'centre of modernity' and an 'economic growth pole'.However, the influx of people overwhelmed the city's infrastructure.On behalf of the city government, which wanted to reduce migration, the City Planning Department changed its policies to restrict the housing market, which led to higher housing prices.Migrants with functioning family networks could counter this through collective agency, either by staying with relatives in Astana or by receiving reverse remittances from their rural relatives to finance an apartment of their own.Using combative agency, migrants without a pronounced social (mostly family) network mitigated high housing costs with measures resulting in precarious housing situations and/ or highly insecure livelihoods (risk of eviction/demolition).However, the 'Astana narrative' is as strong as ever and migrants are still attracted to the city, even in the face of an unfavourable income-cost ratio.
In the countryside of Akmola, some people were constrained in their agency to stay, especially young adults in search of higher education.In response, the Kazakhstani government decentralised higher education facilities, set up special scholarship programmes to attract qualified young professionals as state employees to rural areas, and implemented rural economic stimuli programmes to improve the rural job market.Ideally, these institutional changes will help to develop a positive narrative of rural and small-town economic areas and help dismantle the belief that migration is a precondition for success.As discussed, the Kazakhstani government created various narratives to promote its goals.Nevertheless, it would probably be more effective for the state media to show the life and living conditions of everyday urban citizens in Astana, including precarious housing, the high cost of living and the struggle to make ends meet (for example, the need to take on more than one job) and not only idealised pictures of the modern metropolis.This would give many rural dwellers a more realistic picture of what to expect in Astana.The strong use of family networks for housing and reverse remittances is also a result of the sometimes ill-prepared or even irrational decisions of younger people to relocate.This is also shown in the severe mismatch between the jobs young people often work in and their level of education.Many of the migrants working in menial jobs are overqualified, as our qualitative research revealed.If they do not progress in their career, their substantial investment in their education may not produce the expected return.Thus, providing young adults in the countryside with appropriate information on urban jobs, qualifications, potential income and housing costs may be a more promising approach than simply trying to constrain their agency.This would also lessen the need for reverse remittances to support unsustainable livelihoods in urban areas.
One effect of the Kazakhification policy is that speaking Kazakh has become more important both in the private sector and, even more so, for public employees, who must now master both Kazakh and Russian.Since ethnic Russians often do not speak Kazakh well enough, they are underrepresented in the urban labour market, particularly in public administration.This has had an interesting side effect.Ethnic Russians have a higher intention to stay in rural areas compared to their Kazakh neighbours.If our analysis holds true, in the long term this may even lead to a persistent (re-emerging) Russification of the northern countryside.
By developing a conceptual guide on the basis of New Institutionalism for the deconstruction of migration systems, this study has wider significance that extends beyond Kazakhstan.The analysis of institutions and agency under a cross-school New . Historical institutionalism not only considers time FIGURE 1. CONCEPTUAL FRAMEWORK FOR MIGRATION SYSTEMS BASED ON NEW INSTITUTIONALISM Source: Authors.

TABLE 1 (
Continued)In Kazakhstan, the family network goes beyond the core family and can include around 100 individuals.This extended family network is useful in providing accommodation at expensive city locations(Laszczkowski 2011a, p. 84)or financial support, for example, reverse remittances from the countryside to subsidise city life.According toDietz et al. (2011, p. 21), about 40% of the migrants finance their relocation primarily through family transfers.This in turn strengthens the familial institutions and fosters their continuity.Almost 70% of the household survey respondents stated that they would need prolonged financial support from their families if they moved to a city.

TABLE 2 MIGRATION
Dufhues et al. 2021)).*The'RoadmapforEmployment 2020' (GovReKaz 2016) laid out a plan for the development and support of small-scale businesses in rural areas, for example, through microcredit schemes or farming subsidisation.†Severalgovernmentprogrammes exist to make rural livelihoods more comfortable by improving infrastructure and incomes.‡In the last decade, those migrating to bigger cities such as Astana have often originated from smaller towns.In the Kazakh context, these are considered 'rural'.The government therefore established special economic stimuli programmes for the development of small towns(Bissenova 2017, p. 646).One measure is to relocate state administrative agencies and thus public jobs to small towns and develop economic clusters.In the latter, Soviet-style 'single-industry towns' are transformed into local economic clusters (GovReKaz 2016).

TABLE 2 (
Continued) Rural health and education facilities have undergone notable improvement due to public investment.Nevertheless, they are still falling behind those in cities as this is where economic growth is concentrated.Health and education facilities in Astana are almost always assessed as superior (see TableA2).The same services in the rural areas, however, were considered by only one quarter or half of the household survey respondents respectively as better than average.Furthermore, access to free medical treatment is only granted at the place of registration.For families with children and the sick, this continues to be an incentive to move to and register in larger cities. * Interviews: 8, government official, employment department, Stepnogorsk, 25 May 2016; 9, village mayor, Stepnogorsk district, 26 May 2016; 10, village mayor, Shortandy district, 16 September 2016; 11, village mayor, Akkol district, 19 September 2016.† 'Implementation of the Enbek Programme', Official information source of the Prime Minister of the Republic of Kazakhstan, 5 March 2020, available at: https://primeminister.kz/en/news/reviews/veterinary-requirements-investment-attraction-export-of-productsresults-achieved-in-agriculture-development-of-kazakhstan-in-2019,accessed 23 May 2022; interview 12, town mayor, regional town Pavlodar province, 1 June 2016.‡' V Kazakhstane lyudi chashche pereezshayut iz aulov v goroda', Argumenty i fakty, 12 February 2020, available at: https://kzaif.kz/society/v_kazahstane_lyudi_chashche_pereezzhayut_iz_aulov_v_goroda, accessed 23 May 2022; interviews: 3, village mayor, Stepnogorsk district, 27 May 2016; 11, village mayor, Akkol district, 19 September 2016; 13, town mayor, regional town, Akmola province, 20 September 2016.¶ Interviews: 8, government official, employment department, Stepnogorsk, 25 May 2016; 12, town mayor, Sources: