Private finance integration to affordable housing production: a comparison between Copenhagen and Istanbul

This study analyzes transformations in institutions of affordable housing production connected to neoliberalization, which have impacted affordability rates and income gaps in Copenhagen and Istanbul. Also, different types of affordable housing began to develop, such as modular housing of Almenbolig+ in Copenhagen and housing for very low-income people in Istanbul. Increased affordability rates are the results of the transformations of welfare state-based practices to market-based approaches. The aim of this paper is to analyze and compare how affordable housing institutions have transformed from being financed through public funds to private funding sources since the periods of financial crisis in 2001 and 2008 in a Western and non-Western country. Using the concept of neoliberal localization, this comparative case study employs variation finding tool to analyze the transformations under three causal mechanisms. The first mechanism is state withdrawal of providing finance to affordable housing production; second, collaboration with private sector for financing affordable housing while increasing authority of municipalities. The third mechanism is restructuring institutions to open a base for public-private partnerships. Primary data was provided from interviews, while secondary data was gathered from planning policies, legislations, OECD and country statistics. The outcome of this study reveals information about changes in affordable housing institutions and neoliberal effects on affordable housing production due to contributions of local governments contingent on requirements of private funding.


Introduction
Since the 1980s, production of affordable housing slowed globally, while land and housing prices increased considerably (Rhodes and Mullins 2009). States have contributed to direct support of individual well-being, employment, health, and education in many countries. However, Affordable Housing (AH) production declined in many European countries between 1990 and 2016, see Fig. 1. Reductions in production subsidies for AH given by states was one result of state withdrawal from direct "welfare provision", according to Rhodes and Mullins (2009, p. 112) a first step of embedding private finance in AH production.
The retrenchments in welfare expenses took place not only in Western countries (Morel et al. 2012;Palier and Hay 2017). Also, private finance integration observed with promotions for home ownership and mortgage lending, housing construction booms, and local Public-Private Partnerships (PPPs), which were enabled by neoliberal policies of housing.
As explained by Fox-Rogers and Murphy (2015, p. 42): Associated with the neoliberalization of planning is the growing reliance of local government on the private sector marked by the emergence of Public Private Partnerships (PPPs), quangos, and other forms of collaborative governance arrangements between the public and private sphere. Mayer and Kunkel (2012, p. 18) emphasized that neoliberalization in both Western and non-Western countries have become "destructive" and created "severe social polarization" on the local level. Severe social polarization has been observed with growing income gaps between the richest and poorest in Western and non-Western countries since the 1980s (Mayer and Kunkel 2012). Reducing state support to inhabit or subsidize low income people by deregulation of affordable housing policies have increased spatial inequalities. The issue of deregulation and inequalities are intertwined making it hard to decide which led to the other in "the toleration of rising inequalities or the deregulation that allowed inequalities to rise" (Dorling 2014, p. 126). In other words, the tendency of reducing regulation in housing markets in countries, such as reducing the state support to make Affordable Housing (AH) become accessible is in parallel with increasing gap between the poorest and richest (Dorling 2014). Financial crises in countries are catalysts of change in planning policies to embed neoliberal projects that impose free market rules and lead to increase in income inequalities. Western states respond to financial crises by imposing strict control of public institutions that make markets more flexible to allow entry of private sector (Mayer and Kunkel 2012). Non-Western states have used strategies of Western states to promote private sector by policy transfers (Mayer and Kunkel 2012). For example, IMF provided financial support on the condition of state promotion of PPPs in Turkey and Argentina in financial crisis of 2001, and financial crisis in Mexico of 2007 (Bugra and Candas 2011;Marois 2011;Önis 2006). The common outcome of the process in many countries is the empowerment of private sector, even in AH production. This research answers the question pertaining to how institutions of AH production is restructured to open a base for PPPs after the integration of private finance in two selected cases.
In this study, cases selected from a Western and a non-Western Country involve two different welfare systems keeping one of the lowest (Denmark) income gaps by 0.25 and the highest (Turkey) by 0.398 in 2015 (OECD 2018). AH production as a type of welfare provision in Denmark and Turkey is assumed to act as an obstacle against growing income inequalities. An increase in income inequalities emerged after a gradual privatization process in the mid-80s and became remarkable in the 90s in Turkey (Önis 2011). Even though Turkey has shown a downward trend in income inequality since the mid-2000s, poverty rates are still high (OECD 2011. Denmark has shown a remarkable increase in income inequality between mid-90s and 2000s after Copenhagen city was almost bankrupt (OECD 2011. Income inequalities occurring in both countries are shown in Fig. 2. However, by observing the case studies in this paper, private finance integration has changed AH production in Copenhagen and Istanbul in three ways. First, social expenditure for low income people differed in time. Second, the number and types of AH changed and local Public Private Partnerships (PPP) were used. Third, the affordability rate has become more than 30% due to increases in land prices, construction costs, rents and monthly installments of AH. The term AH is used in this study to explain housing that has been subsidized publicly to meet the needs of a relative community (De Kam et al. 2008). In this study the relative community consists of low-income people and AH provided by state or non-profit institutions (Marsh and Mullins 1998). AH production institutions are defined in this study as institutions responsible for physical production and provision of  Denmark andTurkey (OECD 2011, 2018) 1 3 AH for low income people. AH production institutions are also responsible to specific laws and regulations. Cooperative housing production is not a part of this study.
Neoliberal effects on AH studied considering state support for public housing privatization, encouraging home ownership by mortgage and low production levels of new AH (Dodson 2006;Hananel 2014;Kadi and Musterd 2015). Moreover, Priemus and Boelhouwer (1999) analyzed institutional structure of AH in seven countries in Europe including Denmark in terms of government subsidies, macro housing quotas, financial sources of new production and renovation, governmental guarantees of interest rates and mortgage loan borrowing. In Denmark, the mortgage period for AH construction loans was changed from 35.5 to 50.5 years, which led to reduction in amount of annual installment payments in 1990 (Priemus and Boelhouwer 1999). Additionally, subsidies were provided to lower interest of loans (Priemus and Boelhouwer 1999).
This study builds on recent findings related to the work of Priemus and Boelhouwer (1999) and adds information about the changing roles of state, municipalities and private institutions to finance AH production until 2016. The aim of this paper is to analyze and compare how affordable housing institutions have transformed from being financed through public funds to private funding sources since the periods of financial crisis in 2001 and 2008 in a Western and non-Western country. Even though Denmark and Turkey have different types of welfare provision, tax return systems and different housing market types, institutions of AH production were transformed by similar neoliberal mechanisms.
Transformations emerged with financial crisis in 2001 when financial sources of AH production institutions changed after cutting state budget in Turkey. In Denmark, a private funding institution has started to guarantee majority of mortgage loans for NHAs on behalf of state in 2002. One outcome of this study shows how different types of AH production, Almenbolig+ in Copenhagen and AH for very low-income people in Istanbul, began to develop. During institutional transformations, three mechanisms consisting of events were observed. A historical order of events in each case are analyzed considering neoliberal localization concept to explain institutional transformations.
The first mechanism is state withdrawal of providing finance to AH production; second, collaboration with private sector for financing AH while increasing authority of municipalities. The third mechanism is restructuring institutions to open a base for PPPs. An additional unexpected fourth mechanism, different than neoliberal localization concept, emerged during this study. State came back to the role of welfare provision for financing AH production for very low-income residents in both cases. One of the Turkish ministries subsidizes AH construction to keep rents lower, while Danish state purchases bonds and provides loans from private mortgage banks.

Background
There are four essential differences contingently related to changes in institutions of affordable housing production in the selected cases. These are welfare provision and tax returns, housing market and tenure type; the share of private mortgage loans for home ownership and GDP rates and relative income poverty. First, in Denmark, 20% of income taxes are transferred to local governments to provide welfare to citizens (Goul Andersen 2011). In Turkey, state is the only entity responsible for welfare provision and tax returns. Second, Denmark has an integrated housing market in which the state controls rental amounts for both private and public housing. In Turkey, there is a dual housing market, which does not modify markets. Denmark has a home ownership rate of 50.6%, with an affordable rental housing rate of 22% (Scanlon et al. 2015). Turkey has dual housing market with a home ownership rate of 68%, while affordable owneroccupied housing rate was approximately 10% in 2013 (Coskun 2015). Third, mortgage loan usage rates to purchase houses are different. The rate of residential mortgage debt to GDP ratio is higher in Denmark than in Turkey. The rate is 65% in Denmark, but 5% in Turkey (European Commision 2015, EUROSTAT 2016), shown in Fig. 3 (Tsenkova 2008. Reluctance to use of mortgage loans is due to TOKI's low interest rate was 4% for AH residents making housing purchases from TOKI preferable (TOKI 2011). Since private banks offered mortgage loans with interest rates between 10 and 14% in 2011 (TMB 2018). Fourth, Turkey had the second highest relative income poverty by 17.3% among OECD countries. On the contrary to Turkey, Denmark had the lowest income poverty by 5.5 among OECD countries in 2015 (OECD 2017).
Similar policy changes in institutional transformations such as to promote private sector and local level PPPs for AH production are observed. The first local PPPs for AH productions were established in Copenhagen for the Emaljehaven project in 2008 and in Istanbul for the Sulukule project in 2009. Brenner and Theodore (2002, p. 349) describe "neoliberal localization" process; first limiting public finance for local governments while increasing their authorities and second encouraging adoption of private sector or revenue-sharing methods. A final aspect of this restructuring of institutions is to open a base for PPPs to provide local government services (Brenner and Theodore 2002). Residential mortgage debt to GDP ratio (Tsenkova 2008, p. 25)

Method
This study compares the process of AH institutional transformations and reveals the results considering affordability rates by analyzing the share of rent amounts, or monthly installments of AH units to residential equalized incomes of low-income households in Copenhagen and Istanbul. To compare processes of institutional transformations, successive events are considered during the analysis of institutional transformation that stem from neoliberal influences. The institutional transformations emerged with causeeffect relationships based on housing and planning policies.
The two cases have many variables, referred to as differences, affected by one common mechanism being the integration of private finance in changing institutions in AH production, of which PPP involvement is extracted as one common outcome. This study uses a methodological tool called variation finding comparison to work with systematic differences between cases (Pickvance 2001). Differences can be analyzed under the common feature, a dimension or a concept, called "conceptual equivalence" (Pickvance 2001, p. 24). According to Pickvance (2001) empirical differences need to be considered to design a research study (Pickvance 2001). Mahoney (2000) explains institutional change as a concept of self-reinforcing sequences. According to Mahoney (2000), an unexpected initial event promotes changes in institutions, which is followed by successive events that continue to change the structure of an institution. These successive events generate causal mechanisms that influence the institution to work differently than past actions (Mahoney 2000).
In this paper, an initial event starts with budget cuts in AH production institutions; following events are observed in integration of private finance. The initial event and following events, which create causal mechanisms, are analyzed by a literature review of laws in housing and planning policies. Three causal mechanisms observed based on neoliberal localization concept. The first to limit public finance while increasing authorities of local governments; the second is to encourage private sector or revenue-sharing methods; the third is to restructure institutions to open a base for PPPs (Brenner and Theodore 2002).

Data
The data for this research consists of primary data for analysis of legal documents and interviews. Documents gathered for the Denmark and Turkish cases include municipal plans; laws about rent subsidies to citizens; AH laws, regulations and statistical documents from OECD, see Table 1. The number of AH units produced in Istanbul were obtained from the website of TOKI while Copenhagen Municipality's supplementary planning documents include information about the number of produced AH units. Turkish Statistics Institutute (Turk Istatistik Kurumu, TUIK) documents for Istanbul were used to obtain equalized household income in 2016; the source for Copenhagen was Status på København 2018 to evaluate housing affordability rates.
A rent analysis to calculate the affordability rates in Copenhagen were conducted on 16th of June and in 24th July in 2018. The data for the rent analysis was drawn from the KAB website, a nonprofit housing association, which owns and manages AH mainly located in Copenhagen (Kab-selvbetjening 2018a). Interviews were conducted in Denmark  Tillaeg44 2015, Consolidation Act No. 269, 2017BEK No. 1303 Law no. 4966 (4966/4, 2003) Law no. 5069 (2007) Law no. 5366 (2005) Statistical documents OECD OECD, TUIK, TOKI with a NHA and Copenhagen Municipality. Interviews were also done with representatives of TOKI and local municipality in Istanbul.

Household income in Copenhagen and Istanbul
The population of the metropolitan area of Copenhagen was 580.300 in 2015 (Kommune 2016). The annual income distribution in 2016 reflected that 21% of the population consists of very low income households earning less than 13.000 Euros 1 (Kommune 2018, p. 21). 34% of the population earned less than 26.000 Euros, while 25% of population earned less than 39,000 Euros as middle income holders, as shown in Fig. 4 below (Kommune 2018, p. 21). On the other hand, 18% of the population had high incomes earning less than 52,000 Euros, while very high income households had more than 53,000 Euros (Kommune 2018, p. 21). Istanbul had a population of 14 million in 2015 (Eurostat 2016) and the distribution of annual equivalized household disposable income is ordered by five quintiles by TUIK between 2006 and 2016 for Istanbul (TUIK 2017). A median of very low-income households, consisting of 6.7% of population, was 2712 Euros per year while 10.3% of population was low income households earned 4043 Euros in 2016 (TUIK 2017). 14.1% of households had 5435 Euros who were middle income, 20% of households had 7533 Euros consisting of high income and 49% of households had very high income earning 14,417 Euros annually (TUIK 2017). 1 Euro conversion from DKK and TL was done considering the data of European Central Bank Euro system (EcbEurope, 2016). In 2016, 1 TL was equal to 0,3 Euro, while 1 DKK was equal to 0,13 Euros.
According to Official Gazette of Turkey, a worker receiving social security had a minimum monthly salary of approximately 494 Euros, 5929 Euros annually, in 2016 (OfficialGazette 2015; URL-1 2016). According to TUIK, statistics shown in Fig. 5 below, this example of a worker's wages is claimed to represent a middle-income household in 2016. The living expenses of people at risk of poverty was collaboratively defined by workers' unions to be 525 Euros monthly and 6300 Euros annually (URL-1 2016).
According to given data, people earning low, very low and middle incomes in Istanbul, consisting of 31% of the population, were at risk of poverty in 2016, in contrast to the Official Gazette of Turkey. An umbrella organization of Turkish Chamber of Engineers, Architects and City Planners announced the minimum salary of people who have a university education, as being 990 Euros monthly and 11,880 Euros annually, which was high income in 2016 according to TUIK (TUIK 2016(TUIK , 2017TMMOB 2015).

Analysis of institutional transformations of AH production
This section provides information about the relationships between institutional changes and causal mechanisms in Copenhagen and Istanbul. In Turkey's financial crisis of 2001, the national government began to support urban renewal practices with revenue-sharing methods to promote Istanbul's global competitiveness (Karaman 2008). International Monetary Fund gave funds on the condition that Istanbul is planned to become a global city (Karaman 2008). In Copenhagen, the adoption of European Union policies with strengthened spatial planning since the 1990s, when Copenhagen Municipality was almost bankrupt. European Union policies have been based on diversity, which has bolstered private and localized development by deregulation policies (Galland 2014). Therefore, urban planning changed direction from the stance of equality to liberal by adapting goals to promote privatized local development to become a global city (Galland 2012(Galland , 2014. AH became involved in a series four events, creating financial contexts for the shaping of causal mechanisms of institutional changes in AH production.

Private finance integration in AH production and policy changes in Copenhagen
In Denmark, the production of AH is financed by tenant's deposits of 2%, municipal basic capital between 7 to 14% and credit institution loans with state guarantee from 84 to 94% (Velfaerdsministeriet 2009). State has never directly financed affordable housing construction (Turner 1999). Instead, the Ministry of Housing provided mortgage loans from private banks on behalf of state and payback was guaranteed by state in 1975 (Turner 1999). The amount of state provision of mortgage loans changed from 10% in 1975, reduced to 4% in 1989, then were completely abolished in 1990, Copenhagen Municipality was almost bankrupt, see Table 2 (Velfaerdsministeriet 2009, p. 83). First, in 2002, 100% municipal guarantees for mortgage lending of nonprofit housing associations (NHAs) were lowered (Jensen 2013). Before 2002, municipalities in Denmark were guaranteeing mortgage loans 100% for NHAs. The payback period for the borrowed loans was up to 50 years free of charge until 2000 (Priemus and Boelhouwer 1999;Velfaerdsministeriet 2009). Mortgage loans were given by private mortgage and commercial banks (Boligkontoret 2007) and interest rates were guaranteed by state to be kept low (Tsenkova and Vestergaard 2011). NHAs purchase mortgages by paying for interest rates and bank charges, which was established as equal to 3.4% of building costs produced after 1999 (Scanlon et al. 2015). Mortgage repayment fees became more than the amount borrowed and led state to make profit from selling mortgages, even though the state pays the mortgage costs to private banks (Scanlon et al. 2015). In addition to these changes mentioned, increased land prices led to high AH construction costs and higher rent requirements for AH produced after 1999.
In 2002, the maximum payback period for municipality provided guarantees for mortgage loans via LBF reduced to 30 years (Velfaerdsministeriet 2009, p. 69). Financing rules of AH were changed and stimulated NHAs to produce AH with higher rents to reimburse deficits in later years (Jensen 2013). A private funding institution, Landsbyggefonden started to guarantee 53-71% of mortgage loans for non-profit housing associations on behalf of municipalities (Velfaerdsministeriet 2009, p. 83). Landsbyggefonden is an independent organization established based on a law in 1967 by nonprofit housing associations (LBF 2015). Associations give 50% of their annual income as paybacks of borrowed mortgage loans to Landsbyggefonden (Tsenkova and Vestergaard 2011). Also, 7% of interest free loans were given from municipalities via Landsbyggefonden, as shown in Table 3 above until 2007 (Velfaerdsministeriet 2009, p. 83). Second, the municipality of Copenhagen provided guaranteed mortgage loans that increased to 75% of a property purchase price in 2007 (Velfaerdsministeriet 2009, p. 71). The support of Landsbyggefonden to guarantee mortgage loans for non-profit housing associations on behalf of state was reduced to 25%, while municipal guarantee increased to 75% by Act No. 547, see Table 3 above (Transport 2018, p. 13). Meanwhile, municipal contribution previously provided interest-free loans that increased to 14% in 2007 via LBF, reduced to 7% in 2012, and increased to 10% in 2016 (Finansministeriet 2015, p. 35;Velfaerdsministeriet 2009, p. 71). Then, public housing act of 2013 gave responsibility to Landsbyggefonden to administer loans for NHAs without interest rates until 2018 (LBF 2013;LSF135 2013). Landsbyggefonden provided funds between 2015 and 2018 as guaranteed mortgages for new AH construction done by 25% rule (Transport 2018, p. 13).
Fourth, in 2018 a law came into force consisting of a state-owned system that purchases private bonds to subsidize AH production (LOVN0773 2018; Nationalbanken 2017). To provide loans, the mortgage banks pay a 0.12% loan fee to the state (Transport 2018;Denmark 2017). Law number 0773 led to an increase in construction expenses of NHAs and may lead to a growth in deposit payments of tenants in future. To sum up, financing AH production became more market oriented due to the commitment of purchasing private bonds and loans.

Changing roles of local governments in AH production and policy changes in Copenhagen
A dissimulation of the Metropolitan Council of Copenhagen in 1994 streamlined decisionmaking processes by transferring many duties to city and regional councils (Galland 2014). The local planning duties are under the control of Ministry of Environment (Galland 2014). Also, municipalities had responsibility to determine the quantity of new AH stock in 1997 (Kristensen 2002). The public housing act of 1997 gave a role to city council to control the annual budget of NHAs (Galland 2014). Another change in municipal duties occurred in 2007, which included proving youth courses, health care duties, integration of immigrants to community and creation of local business services (Hansen 2007). In addition to this, the local income taxation kept independent from municipality to municipality on the condition to transfer large amounts to small budget municipalities state (Sellers and Lidström 2007). These changes allowed municipalities to become self-autonomous in relation to urban planning, local infrastructure, and to be powerful actors in AH provision (Galland 2014).
In financial crisis period year of 2008, a local PPP for a new type of AH production called Almenbolig+ with tenure mix strategy appeared in Copenhagen. The Emiljhaven project started with the purchase of the Glud and Marstrand foundation and signing an agreement with Søtoftegaard A/S to develop a tenant mixed housing project on land owned by the foundation (Hansen 2007). The tenant mixed neighborhood consisted of 60 AH units provided by a NHA, 42 cooperative and 83 owner-occupied units (Hansen 2007). The prefabricated unit housing project was completed in 2008 (Hansen 2007). Copenhagen City shares a budget to be used for purchasing land in central locations of Copenhagen to produce AH with 25% rule (Tillaeg44 2015).
The public housing law of 2015 legalized with tenure mix with AH production by reserving 25% of land area (ActNo.221 2015). Three NHAs, KAB, 3B and Domea developed Almenbolig+ houses to be used in tenure mixed projects (Kommun 2014). A first local project by the 25% rule was implemented in the South port area in Teglholmen, Sydhavnen in Copenhagen that consisted of Almenbolig+ units as AH in 2015. The shared budget of city council (Schouenborg 2015, p. 10) is planned to be used to build 10% of AH as small houses on purchased land using 25% rule. Almenbolig+ units are prefabricated (Tillaeg44 2015), used by young people having a maximum rent of €430.48 (Schouenborg 2015, p. 25). The land was purchased by a partnership of NHAs, private and cooperative housing companies. This tenure mixed project includes 6500 square meters AH units produced by this PPP (Turk 2016). Almenbolig+ houses produced as 10% of 6500 square meter area, which were one room houses, consists of approximately 30 square meters rented for 430 Euros.

Private finance integration in AH production and policy changes in Istanbul
The institutional transformation in AH involved in the Istanbul case through four successive events. First in 2001 the financial sources of institution were abolished, and financial resources of TOKI changed (TOKI 2012). Second, TOKI was given authority to implement profit-oriented projects using a revenue-sharing method by law no. 4966 in 2003 to create its own financial resource and cross-subsidize AH production for low income people (4966/4 2003;Palancioglu and Cete 2014). The revenue-sharing is often used in Istanbul by PPPs to capture a rate of profit, usually between 7 and 20% (TOKI 2016) from sales of constructed houses in lieu of TOKI's given public land to private companies with bidding (TOKI 2012). The rate of profit was proposed by the private companies, and TOKI awarded contracts to companies that promised the highest rate of profit sharing. TOKI has mainly produced housing for middle-and high-income groups in Istanbul consisting of villas, apartments, residential towers and gated communities to cross subsidize AH (Karaman 2008). Third, a first historical urban renewal project in Istanbul, started in 2006 by a decision of cabinet of ministers and implementation of law number 5366 (Council 2014). Sulukule urban renewal project was implemented between 2009 and 2012 (Günay et al. 2014). This project consisted of a local partnership among municipality, a private company and TOKI, which included selling 20 AH units for low income residents, who were Roma people (Council 2014: URL-2).
Fourth, social expenditure for low income in AH established by a partnership of TOKI, local governments, and the Ministry of Family and Social Policies by funding from the General Directorate of Social Assistance and Solidarity in 2009 (Ministry 2009). AH production was partly financed with funding since 2010 by law number 5793, even though AH unit production was minimal in Istanbul 2010 (TOKI 2016). The amount of funding became 1.7 times more in 2013 than 2010 (Eryıldız 2017).
Law number 5366 in 2005 gave the authority to cabinet of ministers to approve the sites for urban renewal offered by metropolitan municipalities. Metropolitan municipalities were empowered by implementation authority of urban renewal projects in parallel with given authority to TOKI to implement urban renewal only if metropolitan or local municipalities do not intend to do in 2005. TOKI provided finance to municipalities for urban renewal. Implemented urban renewal projects by TOKI often did not include AH units (Turk and Altes 2010b). In 2007, land access for TOKI was either free or at prices below market by law number 5069 (Palancioglu and Cete 2014). The free land acquisition is requested by the Ministry of the Economy on behalf of TOKI and confirmed by the Prime Minister. As a result, AH has usually been built on public land since 2007 (Palancioglu and Cete 2014).

Result 1: changes in social expenditure for low income people
Social expenditures in Denmark include demand side subsidies consisting of cash benefits as rent and housing assistance, technical education, and integration to the job market that has been managed by local governments in Denmark. Cash benefits given by municipalities were reduced according to Consolidation Act no. 468 (UdbetalingDanmark 2016) and LBKnumber269 (2017) and encouraged finding a part time job. Kontanthjaelpsloft (2016) document reveals the amount of state benefits for a household that consists of a woman aged 30 with two children and a partner (Kontanthjaelpsloft 2016). She received cash assistance in 2017, amounting to 1895 Euros and had rent help amounting to 161 Euros per month for living in AH. In total she had 2056 Euros (Kontanthjaelpsloft 2016). If she works part time at least 6 h a week, she can have more income help, which is 2083 Euros monthly (Kontanthjaelpsloft 2016).
In Turkey, the General Directorate of Social Assistance and Solidarity linked to the Ministry of Family and Social Policies by law no 633 in 2011. The institution helps people giving unemployment salaries who do not have health care and pension benefits linked to welfare system (GDSA 2017). Similar to the Copenhagen case free technical courses are given by local governments to improve skills of low-income people for increasing their employability (Bugra and Candas 2011). According to interviews with representatives of TOKI in November 2016 and information on the General Directorate of Social Assistance website (GDSA 2017), funding is provided by state to AH for very low-income people. The Ministry of Family and Social Policies aims to keep the monthly instalments of tenants in AH very low; to achieve this, the required housing costs of construction are paid by the ministry to TOKI (Ministry 2009). This type of AH has not been produced in urban renewal areas in the historical core of Istanbul, where living expenses and real estate prices have been escalating; they are produced in peripheries.

Result 2: changes in the number of AH in Copenhagen and Istanbul
In Denmark, AH is produced by NHAs. 68% of the household living in AH units are usually single parents with children and 57% of them are single women (Scanlon et al. 2015). In Copenhagen, contrary to general trend of reducing number of AH production in Europe since the 1990s, there is a growing number of AH units produced by non-profit housing associations in Copenhagen as shown in Fig. 6. However, the rent amounts of AH grew 28% in the following years of financial crisis.
TOKI's AH is accessible only to those with certain income limits defined by TOKI. In Istanbul, TOKI completed 100% of 48,683 housing units between 2004 and 2016 (TOKI 2018a). 35% of housing units were AH for very low-income households (TOKI 2018a) and sold to very low and low income households for 15 and 25 year long agreements (TOKI 2014). AH were produced in the city's peripheries referred to as Sile, Tuzla, Hadimköy, Tasoluk, Kayabasi (TOKI 2014, pp. 76,90,91) and Silivri (TOKI 2018b). There was an increase in AH production for very low income between 2007 and 2009, which was followed by a sharp decline in following years, shown in Fig. 7.
23% of housing for high and very high-income people, all produced by revenue sharing, while 42% of them for middle-income people. Housing for middle income people were produced by urban renewal (10%); constructed with (11%) and without (21%) social facilities in Istanbul (TOKI 2018a).

Result 3: changes in land prices, AH rents and affordability
Monthly installments paid by very low income households were a minimum of 90 Euros (300TL) in 2012 in Istanbul (TOKI 2012, p. 25), and increased to 160 Euros (530 TL) (URL-3 2018) in 2016 in Şile (TOKI 2018b), while the installments of Silivri houses are announced as 288 Euros (961 TL) in 2018 (TOKI 2018b). Turk and Altes (2010a) mention that land prices in Istanbul were more than doubled between 2002 and 2008, from 700 to 1900 euros per square meter.
In Copenhagen, an apartment consisting of 85 square meters, constructed before 1980, had a monthly rent of 680 Euros. Figure 8 shows that during 2006 and 2007, AH rents per square meter were between 8 and 10 Euros. During 2010 and 2011, the rent increased by approximately 28% increase to be 268 Euros, which led to rent became 950 Euros.

Rent analysis of Almenbolig+ and traditional AH units in Copenhagen
In 2018, rents of Almenbolig+ houses produced by the PPP in Emiljahaven project in Havnestaden, South Copenhagen are between 1209 and 1716 Euros. The flats consist of three and four rooms with 76 and 108 square meter areas (kab-selvbetjening.dk 2018). The rents per square meter is minimum 14 Euros. Also, traditional AH produced by the same institution before 1960 have 9 Euros, while AH produced in mid-2000 have 13 Euros rents Fig. 8 Changes in rents in AH, which were produced before and after 1980 in Copenhagen in Danish Krone (Kommune 2012, p. 13) per square meter (Kab-selvbetjening 2018c). Almenbolig+ houses are between 100 and 110 square meters including two bedrooms, in Grøndalsvaenge of Copenhagen. Houses have monthly rents between 918 and 1250; which are between 10 and 12 Euros per square meter (Kab-selvbetjening 2018b). AH, which produced before 1960 are rented for 7 Euros and those produced after 2005 have 16 Euros rents per square meter in 2018 (Kab-selvbetjening 2018b).

Discussion of affordability rates in Copenhagen and Istanbul
This study includes calculation of affordability rates for very low-income in 2016. 55% of the population in Copenhagen had low and very low income who earned less than 26 880 Euros per year. Contrary to the Copenhagen case, in the Istanbul case, 69% of population had high and very high income in 2016 (TUIK 2017). Very low-income households earned less than 13 000 Euros annually in Copenhagen, while very low-income households earned less than 2 712 Euros in Istanbul. It is clear from the population statistics that the reason why 64% of TOKI's housing production is provided for middle-and high-income people in Istanbul. Since 2006 in Istanbul, the rate of people earning very high income was never lower than 43.5%. Monthly installments of AH for very low-income people in Istanbul was approximately 160 Euros for recipients earning around 226 Euros, which was 70% of their total income. The official lowest income receiving social security benefits in Istanbul was 494 Euros in 2016 and they paid the same monthly installments to TOKI; therefore, the affordability rate of housing was 32% for official lowest income holders.
In Copenhagen, a household type is taken into consideration consisting of one working person, one unemployed person who receives welfare cash and rent benefit with two children. A very low income household salary was 1083 Euros, had 1895 Euros cash benefit and 161 Euros rent assistance, which was 3139 Euros in total (Kontanthjaelpsloft 2016). If they live in an Almenbolig+ house in Emiljahaven, the required rent is 918 Euros in 2018, at an affordability rate of 29%. If they live in another AH unit, produced in the 1940s, they pay 491 Euros with an affordability rate of 15% and the rent for units produced in 2013 was 1045 Euros while affordability rate was 33%. To sum up, affordability rates of AH produced in the 1940s is 15%, while for those produced in 2013 is 33%.

Conclusion
This study analyzed transformations in institutions of AH production due to private finance integration in Copenhagen and Istanbul, particularly those beginning during financial crisis periods. This research answered how institution of AH production is restructured to open a base for PPPs after the integration of private finance. This study used neoliberal localization concept consisting of three causal mechanisms.
The first mechanism is to limit public finance while increasing authorities of local governments. AH production was 4-10% financed by state, 4-5% by municipalities, 0-8% by Landsbyggefonden, 2-3% by tenants' deposits and 74-94% by private mortgage with 100% municipal mortgage guarantee until 1990. The mortgage costs are paid to private or mortgage banks. In 1990 when Copenhagen Municipality was experiencing financial crisis, the state stopped financial support to AH production. City councils, which are administered by ministry of environment were authorized to control annual budgets of NHA since 1997.
Increased numbers of local authorities emerged with receiving the role from municipalities to define the quantity of AH stock in 1997. Municipalities were authorized to provide local business services, technical courses to adults and to determine tax levies in 2007. In 2015, municipalities had the responsibility to request 25% of land to reserve on land use plans to produce AH. Also, municipalities have right to change the function of the land from industrial or business to residential function, which is applicable by 25% rule. City council provided a budget in 2015 to purchase land in the city center, where land prices were higher.
The second mechanism is collaboration with private sector for financing AH such as to encourage the private sector to enable revenue-sharing methods. In 2002, Landsbyggefonden, started to guarantee mortgage loans in collaboration with the municipality. In 2013, Landsbyggefonden was authorized to administer and guarantee loans for NHAs without interest rates until 2018. Loans were used as funding for new AH construction done by 25% rule between 2015 and 2018. In 2018, state began to guarantee and purchase bonds or loans to finance AH production, while mortgage and commercial banks must pay 0.12% loan fee to state. These evident changes in AH production lead to the appearance of state control being strengthened, and the authority of local governments to have increased since 1999. However, the decisions made by the state and local governments are highly contingent on the need of financial support from private investors. Mortgage costs of NHAs became higher with state purchasing loan and bond fees paid to mortgage lenders according to new regulation in 2018. Indirectly, these fees can lead to high rents and required deposits for tenants.
The third mechanism is restructuring institutions to open a base for PPPs, which was done in Copenhagen by producing tenure mix housing. Almenbolig+ houses with Emiljhaven project, which were developed with tenure mix through a collaboration among NHA, cooperative and a private housing company. Rent analysis has shown that, the rents of AH in Copenhagen had 28% increase 2 years after the financial crisis of 2008. Also, there is a remarkable difference in the rent amount of AH produced in 1940, which has 15% affordability rate, while housing produced after 2005 is 33%, and 29% for Almen-bolig+ houses for very low-income households in Copenhagen.
In Istanbul, the first mechanism of limiting public finance while increasing authorities of local governments, then encouraging revenue sharing methods was experienced in a faster process than Copenhagen. When state budget for TOKI to finance AH production was abolished in financial crisis of 2001, TOKI was given authority to establish its own budget using revenue sharing methods in 2003. Meanwhile, municipalities were equipped with authority to implement urban renewal projects in 2005, while TOKI was authorized to finance and implement projects if municipalities did not. In 2007, TOKI was given free acquisition of land to be used for urban renewal, by approval of Prime Minister. This led to production of AH often done on public land in peripheries cross subsidized through funds from revenue sharing with private companies, which were involved with projects implemented in central locations of Istanbul.
State contribution to AH production for very low-income people is observed in the Istanbul case, similar to Copenhagen. TOKI had financial support from Ministry of Family and Social Policies in 2009 and AH produced for very low-income people with Şile project in 2014. Analysis of income and monthly installments for TOKI houses has shown that affordability rate for very low-income households working in daily jobs is 70%. Also, affordability rate is 32% for people having social security and formal jobs in Istanbul. An acceptable affordability rate is 30%, and if higher, AH is a burden for very low-income households. In Copenhagen, the affordability rate for AH produced in the 1940s is 15% and the rate for housing produced in 2013 is 33%. Also, AH produced as Almenbolig+ housing by PPP has an affordability rate of 29%. This research has shown that AH is a burden for very low-income households in Istanbul. Similarly, for Copenhagen AH produced after 2005 has an affordability rate of 33%, which is a burden for low income households. However, AH produced between first and second world war period have an affordability rate of 15% is not a burden for low income people.
OpenAccess This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.