How Much State and How Much Market? Comparing Social Housing in Berlin and Vienna

This article examines the effect of two different social housing systems on housing-market characteristics and affordability using a case study of Berlin and Vienna. Despite comparable framework conditions, both cities pursued very different social housing policies. While both cities are currently challenged by growing populations and rising house prices, we show that Vienna is more successful in providing affordable housing as it has created effective institutions to attract and retain private actors in this sector. In Berlin, in contrast, the current crisis reveals the lack of institutional capacity to actively pursue housing policy. We argue that different housing market outcomes can be attributed to fundamental differences in the organisation of social housing provision reflecting divergent state-market relations. While in Berlin social housing has been seen as a provisional measure to correct temporary market imbalances, Vienna’s continuous state participation in the housing market has created a framework of balanced rights and responsibilities for social and private actors to achieve social policy goals. Differences in funding arrangements and institutional embeddedness influence not only housing outcomes but also feedback on the state’s capacity to intervene successfully in housing markets.


INTRODUCTION
Housing affordability has re-appeared as a growing problem across the EU, in particular in urban centres where both housing needs and prices are rising (Pittini 2012;Wetzstein 2017). There are various policy instruments to improve affordability, including social housing provision. This article examines the effect of two different social housing systems on affordability and housing market characteristics using a case study of Berlin, Germany and Vienna, Austria. Berlin is taken as an example of a social housing system characterised by a retreat of the state from active housing policies. Public financial support for social housing has fallen over time and capital subsidies which support housing constructionhave been replaced by means-tested revenue subsidies, which support individuals through housing allowances. At the same time, benefits were reduced and restricted to very needy groups. In contrast, in Vienna, continuous state participation in social housing construction has created a comprehensive framework for a permanently rent-controlled segment of the housing market that is available to the majority of the population. Since Berlin's development is characteristic of the market-oriented approach that has prevailed in most European countries since the 1980s (Salvi del Pero et al. 2016;Poggio and Whitehead 2017) this article contributes to a comprehensive analysis of this political change.
Berlin and Vienna are almost ideally suited for this comparison. They share many similarities in terms of their economic and welfare state institutions and also with regard to their very early housing policies. However, as far as more recent social housing policies are concerned, the two cities have pursued very different policies, reflecting different relations between state and housing markets in otherwise very similar settings. Our comparison rests on the logic of a most similar systems design (Przeworski and Teune 1970) where the variable of interest varies in very similar contexts.
The article follows a neo-institutionalist view that regards institutional arrangements as key factors in shaping market outcomes (Fligstein 2001(Fligstein , 2015. In this context, states are crucial for the formation and ongoing stability of markets as they provide reliable and stable conditions under which exchange can take place. By defining what is legal and what is illegal in a housing market, the state determines the relationship between property owners and everybody else in society, with varying distributional effects. A state's capacity for market intervention, regulation, and mediation depends on the particular history of the state and how organised groups could and can enforce their interests (Fligstein 2001, 42). State-market relations may change over time and some states have greater capacity for intervention than others. An in-depth approach of two cities is chosen in order to compare how differences in the organisation of social housing provision reflect divergent state-market relations and how they affect market developments.
A comparison of the programmes reveals that both governments developed different kinds of capacities to intervene in their housing markets. The article concludes that different market outcomes can also be attributed to fundamental differences in funding arrangements, institutional embeddedness and the long-term orientation of the cities' social housing policies. Vienna's approach of promoting permanent affordable housing is more successful in that it enables the city to actively pursue housing policy goals. Berlin's approach of temporary state intervention, on the contrary, has meant that the effects of earlier subsidies have been almost completely lost and the city today has neither the institutional capacity nor the financial means to pursue an active housing policy.
The article starts with a brief overview of the institutional commonalities between the two countries followed by an analysis of major differences in their housing policies. Section three discusses housing market outcomes and relates these to social housing programmes. The article concludes with a discussion of the effectiveness of both systems and asks to what extent Berlin could adopt policy measures from Vienna.

Similarities Between Berlin and Vienna
Berlin and Vienna are the capitals of two countries that are usually grouped together in international classifications, e.g. Esping- Andersen (1990), Hall and Soskice (2001), or Amable (2003). Consequently, both countries display many similarities in their economic and welfare state institutions. Germany and Austria also share similarities in their housing market characteristics such as well-developed rental markets, low homeownership rates and conservative lending standards which are considered to be reasons for the lasting housing market stability in recent decades (Schneider and Wagner 2015). In contrast to most European countries and the US, house prices in Germany and Austria did not start to rise until the Great Financial Crisis in 2008 (de Boer and Bitetti 2014). Recent price increases in urban areas in both countries have been driven by demographic changes (primarily immigration and an increase in the number of single households) as well as by increased demand for safe assets and low financing costs, which made real estate in both countries an investment for global and local investors (de Boer and Bitetti 2014). As a result, both countries face rising housing costs in urban centres and acute housing shortages in the low-and middle-income segments. The two cities are also very close in other aspects that can influence rent prices (Egner and Grabietz 2018): net hourly wages, cost of living, population density and structure are nearly identical in both places (MA23 2019).

Differences in Housing Policy
Historically, social housing policies in Berlin and Vienna also shared many similarities: Both cities had a municipal housing sector before World War II as well as a tax-privileged limited-profit housing segment, and both have tackled severe housing shortages after the War with broadly targeted capital subsidies. Despite these parallels, the current affordability crisis has highlighted the profound differences in current housing policies. The systems are organised very differently and reflect diverging views on the desirability of market mechanisms in the housing sector. While Vienna is a prime example of the Austrian tradition of supporting a large and permanent limited-profit sector, Berlin embodies Germany's tradition of more limited state intervention in the housing market; the latter typifies the concept of the social market economy where regulatory policies aim to provide a legal and institutional framework for market processes and to correct market deficiencies without distorting competition (Wagner 1995). In particular, housing-market interventions are seen as a temporary measure to overcome acute shortages until markets could do their work in balancing supply and demand (Weber 2018). This is not the case in Austria, where policies aim at maintaining a partly de-commodified sector to secure the provision of affordable housing (Brunnauer et al. 2019). Consequently, the national regulatory frameworks in which both cities operate reflect these fundamentally different views. The following section describes basic features of the German and the Austrian social housing systems and takes a closer look at various dimensions of the programme structures: subsidy and funding modalities, targeting, and the institutional embeddedness of the programmes.

Basic Ideas Underpinning Social Housing in Germany
In Germany, the term social housing typically refers to a complex system of public intervention in housing policy. There is no universal model of social housing that applies across all of Germany as the states and municipalities have great discretion over their housing policy. In addition, funding schemes have changed over time, resulting in a wide variety of funding arrangements. However, the main feature of social housing in Germany is that a public subsidy is delivered to any kind of housing provider in exchange for letting the dwelling to eligible households under regulated conditions for a fixed period of time. 1 Providers of publicly subsidised housing include municipal housing companies and cooperatives as well as private and commercial developers or investors. Technically, all housing providers are considered market actors, although municipal housing companies often have to act according to local policies (Housing Europe 2010).
The funding can be described as a form of public-private partnership. Public subsidies can take different forms, but the basic idea is that public subsidies cover the gap between the reduced rentwhich is charged to eligible tenantsand the cost-based rentwhich is determined by production cost and funding modalities. The subsidy decreases progressively over time and the reduced rent rises accordingly until it reaches the cost-based rent. Typically, the amortisation period is reached after 20-40 years. After that, the subsidised rent has reached the market level (sometimes beyond that if building costs were higher), and the dwelling is transferred to the private segment of the market. A direct consequence of the time-limited social obligation is the built-in loss of social housing at the end of the amortisation period which leads to a steady decline in the stock of social housing if losses are not replaced by new construction. As the German state gradually withdrew from social housing from the 1980s onwards, the stock began to decline.
Traditionally, the federal government was the main funder of social housing. In the 1950s, spending on social housing accounted for 4 per cent of the entire federal budget (Huebl and Moeller 1997). At that time, 86 per cent of all new-built housing units were publicly subsidised (KfW 2008). Since the 1980s, public subsidies have been cut on the grounds that adequate supply and generally high housing standards had made public interventions superfluous except for vulnerable groups (Droste and Knorr-Siedow 2007;Andersen and Munk 1994). The programme was abandoned in 2001 and exchanged for a diversified system of federal and state-funded measures. Since 2006, social housing has been the responsibility of the federal states, and these have implemented very different programmes and funding schemes. The federal government's social housing policy and corresponding financial aid ceased to exist since then. Transitional financial compensation will be available until further notice until 2021 (BMI 2018). This substantial reduction in capital subsidies was accompanied by increases in expenditure on individual benefits (KfW 2008; Deutscher Bundestag 2017c). One motivation was to limit housing support exclusively to groups without adequate housing access. While earlier social housing covered 70 per cent of all households, since 2001 it has been directed exclusively at particularly vulnerable groups (Housing construction law, WoFG). The sharp rise in the shortage of affordable housing in urban centres has led some German statesincluding Berlinto relaunch their building programmes but completions are well below the number of units losing their social housing status (Deutscher Bundestag 2017b, 2017c. This systematic decline has been further accelerated by the fact that housebuilders who have previously received financial subsidies can repay all government-sponsored loans prematurely and free themselves from any social obligations. Given current low interest-rate levels, the majority of private housebuilders in Berlin have availed of this opportunity and refinanced at lower borrowing costs (GEWOS 2016). Depending on the funding arrangements, those housing units lose their status as social housing immediately or after a period of ten years, when they can be offered on the private market.
The shrinkage in housing stock was drastically accelerated by a fundamental change in tax law. The abolition of the special legislation for limited-profit housing associations (Wohnungsgemeinnuetzigkeit) in 1990 made it legal for those associations to sell their rent-controlled housing stock and also ended the tax privilege associated with their business model. This change in law had far-reaching implications as it ended the tax incentive for permanently price-reduced housing. Until then, these associations were required to offer permanent cost-based rents, and profits had to be reinvested in housing. These requirements made trading in those houses unattractive for short-term investors; the change in law prompted the sale of numerous inexpensive apartments to commercial investors. Technically, social housing and the tax privilege were two different things but both relate to government support for affordable housing and have often been interlinked in the past. In any case, the abolition paved the way for subsequent housing stock privatisations and resulted in the loss of a fundamental housing-policy instrument.
This state's withdrawal from the provision of affordable housing reflects a change in the function of housing policyfrom a universal social policy towards targeted policy for marginal groups only (Heinelt 2004)-and illustrates a profound change in social orientation of housing policy. A more or less balanced market has contributed to transforming housing policy into housing-market policy, where tenancy law and individual subsidies have come to the fore. The remaining instruments are no longer aimed at the general public but rather at risk groups or locally unbalanced housing markets (Heinelt 2004). The massive capital subsidies after the War were justified as emergency measures but never conceived as permanent market intervention (Leidner 1981). The abolition of institutional limited-profit housing associations in 1990 was also justified on the basis of its market-distorting effects, the burden on taxpayers and the argument that there was no longer any need for such intervention (Hunger 2017;Holm, Horlitz, and Jensen 2017). The policy pursued can be summarised as a neoliberal shift in housing policy characterised by cuts in spending on supply subsidies, liberalisation through the abolition of the special legislation for limited-profit housing, privatisation of formerly public housing and a shift from a universal housing policy to a targeted social policy (Holm, Horlitz, and Jensen 2017;Egner 2014;Schoenig 2018). This shift was also possible because housing policy was not seen as a direct social policy but rather as market correction. This is in stark contrast to the development in Austria.

Basic Ideas Underpinning Social Housing in Austria
In Austria, like in Germany, public and municipal housing has a long tradition going back to the early twentieth century. While a significant share of the current housing stock was developed in this time, the current institutional framework of social housing originated after World War II, whenlike in Germanythe reconstruction of war-damaged housing was considered a nationwide challenge. Public support through generous object subsidies was essential for the provision of urgently needed homes. Austrian housing policy at the outset was under the authority of the Federal government but over the last 30 years the responsibility has gradually shifted to the regions. Today, the regions exclusively implement housing policy but the state still holds the competence regarding tenancy and tax law.
Unlike Germany, the limited-profit sector constitutes a key component of the Austrian social housing system. Vienna's social housing is generally open to public and private housebuilders and consists of three elements: public housing, provided by the municipality; housing provided by limited-profit housing associations (LPHAs); and different private providers receiving public subsidies. The share of the latter plays a minor but growing role in housing provision. From 2008 to 2018 about 4500 flats were built by those private housebuilders compared to 35,000 flats built by providers that operate under the Limited-Profit-Housing Act (MA50 2019). Limited-profit providers have restricted business activities and assets that are bound to common interest. For-profit providers that use subsidies have to follow most of the rules as well (rents restricted to the 'cost rent', adequate interest rate on their own capital, occupancy rights for one third of the housing units reserved by the municipality) during the amortisation period of typically 35 years. However, after that period they are freed from any regulations, while for limited-profit providers those rules continue to apply.
LPHAs date back to the early twentieth century and have continuously gained importance. Their core function is to establish a long-term social housing stock at below market rents directed at large parts of the population (Mundt and Amann 2009). LPHAs can be owned by local or regional public bodies, trade unions, religious institutions, banks, cooperatives, companies or private persons (Mundt and Springler 2016). They are all regulated by the Limited-Profit Housing Act and controlled and represented by the same umbrella organisation and by the regional governments. Limitedprofit providers are exempt from corporate income tax. In exchange, they are obliged to re-invest any profits in new housing, refurbishment or land purchase. Interest paid on equity to the owners and shareholders is limited to 3.5 per cent of registered capital. The law also prescribes how to calculate rents by following the cost coverage principle in combination with rent limitations defined by the subsidy scheme. Rents must include an amount dedicated to repairs and long-term maintenance. As a result, housing units provided by LPHAs are often in better condition and of higher standards than those owned by other providers (Tancsits 2008;Amann and Jurasszovich 2016).
All these regulations foster a long-term perspective that is strengthened by the fact that housing associations in general manage the houses they have built (Amann and Mundt 2005). LPHAs have grown more significant due to their high levels of construction output. Their obligation to build and reinvest any additional profit in housing leads to a continuous, demand-driven investment independent of economic cycles. In addition, the efficient operating of the LPHA sector contributes to lower construction costs. A study that compared LPHAs and commercial developers found that for newly built rental buildings of comparable quality, construction costs for LPHAs were on average 20 per cent lower (Amann and Goetzl 2005) due to their strong market position for construction materials and services.

Funding Arrangements
This section compares the particular funding modalities of social housing programmes in both cities. In West Berlin, until the late 1960s, housebuilders were supported with low-interest building loans (usually over a period of forty years), which on average covered 50 per cent of construction costs. For the state, this subsidy system meant high costs at the beginning and an exposure to the interest rate risk. In 1969, West Berlin switched to a different funding arrangement. Housebuilders themselves raised private bank loans and received continuous reimbursements from the state to repay their private bank loans. The rationale behind this new arrangement was that the state avoided high expenditures at the beginning and dispersed them over a long time period instead. Yet, in the long run, costs were higher and expenditures for new projects had to be added to the existing costs. In 2001, public debt from social housing claims had increased so drastically that Berlin decided to abandon the social housing programme, as it was no longer able to finance its annual liabilities. As a result, all buildings not refinanced until then immediately lost their status as social housing. More importantly, budgetary pressures prompted the city to sell thousands of municipal buildings to private investors. Another consequence of this subsidy type was that there was no incentive for investors to minimise build costs, as they received funds according to the costs incurred. This effect was reinforced by the use of social housing as taxsaving depreciation object which incentivised increased construction costs and partly explains Berlin's high cost rents (GEWOS 2016). When assessing these funding arrangements, it should be noted that West Berlin's position as an island surrounded on all sides by East Germany until 1989 meant that investing in housing was considered a risky business which is why conditions were made particularly attractive for investors. Still, the basic logic of the subsidy applies to the whole of Germany.
In Vienna, building loans have a maximum amount and are granted irrespective of the costs incurred. This creates incentives to keep costs low and to increase efficiency. Although housing subsidies do not cover the major part of construction costs, they are still necessary for many builders. As in Berlin, the basic idea is that public subsidies are used to reduce and stabilise rents during the amortisation period of typically 35 years. However, unlike in Berlin, LPHAs have to maintain rent regulation forever. Since credit costs are considered when calculating the cost rent, rents have to be lowered to €3.86/m 2 after commercial and public loans are paid back. Here, a major difference to the Berlin system appears: while in Berlin housing units are offered at market level, in Vienna LPHA homes have to be offered at below-market rents after the end of amortisation period.
A Viennese peculiarity is the tenant contribution in LPHA buildings, which covers up to 12.5 per cent of the construction costs and 100 per cent of land costs limited to €300/m 2 . The tenant contribution works as a quasi-loan and when moving out, the tenant gets back the initial sum depreciated by one percent each year. While this contribution was initially a way for new cooperatives to raise capital, it has become a common part of LPHA funding. Its growing use results from increasing land prices and high quality standards (Amann and Jurasszovich 2016) and has been criticised for being a barrier to access LPHA housing for low-income households (Kadi 2015;Amann and Jurasszovich 2016;Reinprecht 2017). To reduce affordability problems, a means-tested public loan was introduced. Another attempt to ease the financial burden was the introduction of so-called SMART flats in 2012 which receive a higher share of public subsidies per m 2 and require reduced tenant equity. In return, those homes have less floor space and are more economically equipped. In an attempt to balance the opposing aims of providing high quality housing for many households without excluding poorer households, one third of all new social housing has to be provided as SMART home.

Programme Funding
Differences can also be seen in the funding of subsidy programmes. Traditionally, in Berlin, most of the subsidies came from federal funds. Since the expiry of the federal programme in 2006, Berlin has received compensation payments and has provided a large part of the funds itself. Whether federal funds will continue to be paid after 2021 is still unclear, as is the question of what resources Berlin can provide in times of tight budgets.
In Vienna, subsidies in social housing are mainly financed by a state-level 'housing tax' collected as a payroll tax (one per cent of all wages, contributed equally by employers and employees) and then passed on to the regions. Another important source for housing expenditure comes from reflows from former loans into the housing budget. For 2018, the gap between revenues and expenditures was financed from the city's budget and corresponds to 0.5 per cent of the city's GDP (City of Vienna 2019a, 2019b.

Targeting
There are also differences in who benefits from the subsidies. Germany's shift of the programme's focus from the majority of the population towards low-income households is in stark contrast to Austria's orientation towards the majority of the population. In Berlin, one-person households are eligible for social housing when they earn below 84 per cent of the median income. For a family with two children the limit is 64 per cent of the median income. Based on these income limits, 831,100 households or 41.5 per cent of all households in Berlin are eligible for social housing (IBB 2019) while the social housing stock amounts to around 100,000 housing units. 2 Income limits in Vienna are more generous, covering around 80 per cent of all citizens (Salvi del Pero et al. 2016). Compared to Berlin, Vienna's income limits are 2-2.5 times higher at around 200 per cent of the median income (wohnservice-wien 2019; Statistik Austria 2019). These high thresholds are at the heart of Vienna's social housing approach, which aims to prevent social housing from being marginalised and to secure majority support in the population. Broad targeting is also part of a wider strategy to influence housing market developments. Various authors (Mundt 2018;Kadi 2015) agree that moderate average rent levels can be attributed to lower rents in the social housing stock. At the same time, there is always the risk of households with higher incomes crowding out more vulnerable households. Finding a balance between social diversity and social targeting is one of the main challenges of Vienna's social housing system. To guarantee a fair allocation, the city developed a complex allocation scheme with the aim to keep broad access without disadvantaging more vulnerable groups (Foerster 2013).

Institutional Embeddedness of Housing Policy
Another important difference between the two cities is the institutional embedding of housing policy in a supportive institutional framework. The Austrian tax law, through the federal housing tax and the tax exemption for limited-housing associations, is consistent with housing-policy requirements. The same applies to the legal system. Recently, the Supreme Court explicitly recognised that it is in the public interest to guarantee the affordability of housing, in particular for lower-income groups and to reduce price perceptibly (City of Vienna 2018). In response, Vienna changed its building code in 2018, attempting to limit land speculation and land-price increases, which istogether with declining land availabilitythe biggest challenge for Vienna's housing policy. The new building code will include 'areas for subsidised housing', in which two thirds of all new apartments must be built as rent-controlled housing as a condition for rezoning (wien.gv.at 2018). The city also has a longer tradition of purchasing land at affordable prices long before turning it into developed land. Despite declining reserves, Vienna is still the owner of larger, undeveloped plots of land, which can be used for urban planning. Since 1995, developer competitions have to be held if the land in question belongs to the city or if new subsidised projects with more than 500 apartments are created. This gives the city a tool to promote socially and ecologically sustainable buildings. Coordination between various institutions contributes to the success of the system and contrast with Berlin. Although Berlin resumed its social housing programme in 2014, there is no specific support from other institutions. For example, a recent Federal Court ruling decided that permanent rent controls are incompatible with property rights. 3 A notable exception is Berlin's 2014 cooperative building development model. This regulation stipulates that 30 per cent of all new buildings must be social housing if the land used is newly converted building land. Between 2014 and 2018, 1933 units were built under this resolution (Berlin Parliament 2019b). Yet despite these modest numbers, the ruling is a first step towards linking land policy and social housing construction. However, the ruling does not reach inner-city areas as it only applies when land is rezoned for development. In inner zones which are not covered by a qualifying land-use plan, developments have to be permitted as long as the building blends with the characteristics of its immediate environment (Section 34 Building code BauGB). Here, public investors can implement their high-priced building projects without the option of public involvement. Vienna's building code also provides for municipal influence only in the case of rezoning (Section 1a Vienna Building code).
Thus far, this section has revealed fundamental differences in the funding arrangements, targeting and institutional embeddedness of social housing programmes. While in Berlin, social housing was never meant to be permanent, Vienna has pursued a longterm perspective and created a diverse and interdependent social housing system, defined by key actors and legally and politically embedded policy instruments.
Such a consistent institutional embedding has also been possible because Vienna has been under social democratic government since 1919, only interrupted in the Second World War. Still, it must also be taken into account that many regulations (e. g. tax law) are decided at the federal level. The broad political support for the limited-profit regulation is also attributable to the fact that many non-profit developers were founded by municipalities (governed by both, conservative ÖVP and social democratic SPÖ), trade unions and also churches (GBV 2016).

Outcomes
While the previous section compared housing policies, this section compared housing market outcomes. Currently, both cities face the same challenge of rapid population growth and growing prices after decades of stagnation. As depicted in Table 1, house prices are rising much faster than population numbers or average incomes in both cities, with relative price increases over the last ten years being stronger in Berlin than in Vienna. Rent increases in social housing were more moderate than in the private sector in both cities. However, one should note that the social sector comprises over 40 per cent of all households in Vienna but below 6 per cent in Berlin. Although in Berlin between 1953 and 1995, more than 521,000 subsidised dwellings were constructed, less than 20 per cent of them are still in this segment (Berlin Parliament 2018). Back then, social housing accounted for almost 85 per cent of the city's newly built housing stock (Huebl and Moeller 1997;IBB 2002). Between 1990 and 2018 alone, this stock decreased by 70 per cent (IBB 2002(IBB , 2018. In Vienna, in contrast, the share increased from 35 per cent to 43 per cent between 1991 and 2018 (Herdin et al. 2010;Statistik Austria 2018). Of this permanently rent-controlled housing stock, 23 per cent is owned by the municipality and 20.1 per cent by LPHAs (Statistik Austria 2018). While the number of municipal housing units has been stable since the 1990s and the municipality concentrates primarily on managing and renovating existing housing, the housing stock of LPHAs is rising steadily.
What is more, price differences between the two segments are bigger in Vienna than in Berlin: In Vienna, social tenants pay 3.87 €/m 2 in municipal housing and 4.84 €/m 2 in LPHA housing compared to the citywide average of 5.27 €/m 2 for sitting tenants (Tockner 2017). However, in Berlin an odd phenomenon has emerged: rents in social housing are sometimes higher than those in the private rental market. More than 65 per cent of all social housing tenants have to pay rents that are above the average private-market rent (IBB 2018(IBB , 2019, so that on average there is no difference in the rent level in both segments. Between 2003 and 2014, when the housing market was relatively relaxed, rents in social housing rose even faster than in the private segment (GEWOS 2016). This is the result of cost-based rent as described earlier: rents in social housing are not based on affordability considerations but on the costs incurred during construction, including interest on loans and guaranteed returns on capital employed. In addition, social housing rents can vary widely, up to 100per cent between different units (GEWOS 2016), reflecting the way cost-oriented rents are calculated. As a result, social housing is regularly too expensive for welfare recipients  and households below the poverty line. Due to rapid population growth, both cities face the great challenge of having to build new housing quickly. The estimated demand for new buildings in both cities accounts for more than 10per cent of the existing stock. In Berlin, there is a projected total need for new construction of 194,000 apartments by 2030 (IBB 2019). A study by  analyses housing demand by income and finds that 130,000 flats for households with income below 80 per cent of the median income are already lacking in the stock, without accounting for future increases. In Vienna, the absolute housing demand is estimated to be between 90,000 and 115,000 units by 2030 (Vienna City Administration 2019) to supply the growing population. The current backlog demand is estimated at 8000 units (Vienna City Administration 2019). The figures show that both cities face great challenges, with the additional demand for apartments in the lower price segment in Berlin being even more urgent than in Vienna.
As illustrated in Table 2, Vienna is currently more successful in providing new housing in relative terms. However, as argued above, the current challenge is not only to build new buildings but above all in the low-cost segment. Since commercial developers build almost exclusively for the pricier segments under current conditions, public programmes are expected to step in. After Berlin restarted its programme in 2014, 1754 housing units were built until 2018 using these subsidies. In the same period, Vienna built almost 20,000 social housing units despite the city being about half the size of Berlin. Subsidy costs per unit in Vienna amount to 510-850/m 2 or €40-60,000 per unit. The corresponding figure for Berlin is €1300/m 2 or max €90,000 per unit of which 25 per cent or up to €22,500 are provided as a redemption-free grant. Vienna not only promotes more output for less money but spends it more efficiently, as subsidies reduce rents for an unlimited time, at least when LPHAs build the homes.
In Berlin, rent controls in the newly designed programme are again only temporary. Although the subsidy is granted as a building loan and sets an expenditure limit, it does not solve the fundamental problem of the time limit, as it only reduces the rent for thirty years. Another problem is that units built with these subsidies are too expensive for households with an income below the poverty line. Initial rents start between 6.50 €/m 2 and 8.00€/m 2 and rise annually by 0.20€/m 2 . A paper by  showed that even initial rents are mostly unaffordable for poor households. An equally important point is that private housing developers have no interest in taking advantage of these subsidies (BMV 2014). They find enough profitable opportunities in acquiring existing stock or building for high-income households. Low interest rates and short-term profit opportunities render a long-term commitment to low rents financially unattractive. In comparison to earlier programmes, nowadays municipal housing associations are the main recipients of those subsidies (Deutscher Bundestag 2017c) but they alone can by no means deliver the required output (GEWOS 2016). With around 320,000 apartments, their share accounts for around 17 percent of Berlin's housing stock (IBB 2018). This contrasts with an estimated demand for almost 200,000 apartments by 2030. In addition, Berlin's housing associations are faced with the dilemma that on the one hand they have to offer inexpensive housing, but on the other they are expected to construct new buildings that are affordable and ecologically sustainable despite high land prices (Berlin Senate 2019). Thus far, a direct comparison of the rental stock in both cities reveals that Vienna has a significantly larger social housing stock and that rents thereunlike in Berlinare significantly lower than in the private sector. The lower rents apply for the lifetime of the building and tenants have long-term security. This availability of affordable housing is reflected in the different housing costs burden: The share of households which are overburdened by their housing costs (i.e who spend more than 40 per cent of their disposable income on housing) differs with 7.1 per cent in Austria and 14.5 per cent in Germany (unfortunately, no comparable, city-level data is available) according to Eurostat (2018).
Comparing the current social housing output illustrates the cities' different capacities to meet the challenge of providing large quantities of affordable housing. If we compare Berlin's current output to demand, it becomes clear that quantities are not even sufficient to replace the annual decline in the stock.  estimate that if the current funding volume were maintained, it would take almost 200 years to close the current gap in affordable housing provision. Even though the current support is helping to increase the stock of affordable housing, the programme does not seem fit to fill Berlin's supply gap. In terms of affordability it is also not an option for poor households. The direct comparison with Vienna's steady output of permanently affordable housing that is added annually to a large stock illustrates the different effects of the two policies.
In short, this sub-section has shown that Vienna's housing market is more balanced, mainly because the demand for low-cost housing meets a larger supply. Vienna is more successful in promoting permanently affordable housing and also more likely to do so in the future. The city has not only a significant publicly-regulated housing stock but possesses effective institutions to attract and retain private actors in this segment. In Berlin, in contrast, it is plausible that the supply of affordable housing will continue to decline, as there is not significant capacity to replace losses in stock. Having illustrated differing housing market outcomes, the final part of this section asks how much these differences are driven by policy.
How did Housing Policy Lead to Different Outcomes? Considering that the framework conditions for the housing markets in both cities are comparable, it seems reasonable to assume that different outcomes can also be attributed different housing policies. An overview of the most important differences shows that instruments are difficult to separate and often interdependent.
Transitory subsidies vs. efficient and sustainable use of funds. The most fundamental difference in the affordable segment of each housing market relates to the temporary use of the subsidies in Berlin as opposed to the long-term effect in Vienna. Vienna's affordable housing stock creates funds that can be used for further housing provision and maintenance. The city aims to continuously increase the share of low-income housing through object subsidies. These subsidies also support long-term cooperation with private actors who accept limited returns and have both the know-how and the capacity to build large quantities of apartments. The combination of legal regulation and capital subsidies creates an incentive structure and gives Vienna the opportunity to actively pursue social policy and urban planning. In Berlin, in contrast, developments over the past decades show two fundamental weaknesses in the programme's logic: First, as soon as the state reduces its financial support, the stock decreases. And second, it is difficult to promote social housing where it is most needed: in growing markets, where investments in other segments are more profitable.
Ad-hoc measures vs. continuity and long-term orientation. The Vienna example shows that sustainable financing and continuous and forward-looking policy are mutually dependent. Especially in residential construction, with its long time horizons, it is necessary to wait until investments pay off. Political decisions that make this long-term approach possible are, again, based on the assumption that housing needs cannot be met by markets aloneas is the case in Berlin. Instead, Vienna's policy reflects the view that a permanent segment operating independently of market principles is needed. Steady housing tax revenues enable continuous demand-oriented investments independent of economic cycles. In addition to sustainable financing, the long-term orientation also has an impact on planning security for investors. Despite regulation, there are enough private investors for social housing construction. In contrast to this coherent policy, Berlin's policy has been more of a patchwork of policy measures, increasingly leading to often-costly ad-hoc measures in order to avoid the immediate collapse of the system or to alleviate severe hardship caused by part of the regulation.
Insulated policies vs. institutional embeddedness. Vienna's long-term orientation is complemented by an institutional embedding that supports the achievement of housing policy goals. Further support comes from tax policy, planning law and land policy and enables proactive action. Berlin's isolated programme, on the other hand, serves more to correct undesirable market developments than to actively control them. Finally, Vienna's broad targeting not only ensures political support but also increases the effectiveness of policy measures.
Overall, a direct comparison of social housing programmes in both cities suggests that the different housing outcomes can be attributed to different policies. Vienna's housing model has proven itself to be a long-term strategy that provides high-quality housing at (mostly) affordable prices. State regulation created a framework of wellbalanced rights and obligations for the social and private sectors. To an increasing degree it has developed market-based methods of finance and remained competitive with the private market. It is this position between state and market that largely accounts for the efficiency of Vienna's housing policy (Mundt 2018) and gives the city the capacity to actively pursue housing-policy goals and adapt them to changing conditions. Berlin, by contrast, has hardly any instruments for maintaining and expanding its social housing stock. Private investors are not interested in current programmes and there is little left of the spending of the last decades. Other important housing market regulations such as tenancy law are determined at the federal level. Because its current design failed to shield tenants from massive rent increases Berlin has decided to introduce a rent cap that 'freezes' the rent at the level of 18 June 2019. This unilateral move by the city of Berlin is an attempt to deviate from existing federal law. The future will show whether this vehemently opposed move (Tagesspiegel 2020) is legally tenable. However, the Viennese example shows, however, that social housing construction can be organised differently. A look at the various dimensions reveals that social housing cannot be understood as a single financing instrument. Rather, the aim is to create structures in the medium to long term that enable the municipality to pursue their policy goals (Ludwig 2017).
Current developments in Berlin suggest that it is necessary to consider a fundamental reorientation of housing policy that justifies interventions not only in times of crisis. The prerequisite for this is a functional rethinking of the question of what housing policy should achieve, followed by an examination of the forms of funding and the institutional organisation of social housing provision (Schoenig 2018). The neo-institutional approach can help to understand that markets have histories, dominant groups, dominant ideas and trajectories (Fligstein 2015, 240). At present, the German housing market is in a state of flux as the existing rules are being challenged and rewritten and there is room for negotiation on how much state and how much market should prevail. This current study compared Berlin's and Vienna's social housing policy and concluded that Vienna's support of a permanently decommodified housing sector helps to promote the provision of affordable housing. In Berlin, in contrast, temporary measures to correct market imbalances in the provision of affordable housing have failed.
The question of to what extent institutions can be transferred to another environment always leads to the conclusion that the respective situation has grown historically and is institutionally embedded. Yet in the case of Berlin and Vienna there are numerous similarities that justify the question of transferability. If Berlin really wants to actively maintain a growing stock of affordable housing, its current funding practice must be changed so that expenditures will have a lasting effect. Permanent ties can be secured either by a land register or by a permanent company obligation, as in Vienna. The legislation on limited-profit housing, repealed in 1990, followed the latter and secured fixed costbased rent limits in return for tax advantages for associated companies, while imposing profit limits and reinvestment requirements. In contrast to provisional capital subsidies this legislation was designed as a permanent arrangement. As seen in Vienna, tax exemptions on public interest is still a successful instrument, even in growing markets, as it offers sufficient incentives for private investors. Yet, a re-introduction of such an exemption would require an amendment of the German federal law. In 2017, a proposal to amend the law to re-introduce special legislation for limited-profit housing was rejected by the German parliament, indicating that currently there is no political majority for such a reform (Deutscher Bundestag 2017a). A paper by Holm, Horlitz, and Jensen (2017) shows possible ways in which long-term commitments can still be achieved via the land register without amending federal legislation. For example, a mix of financing instruments and contractual elements such as ground-rent calculations can be tied to a secured rent level. Although these methods are increasingly used (BMI 2019), they are still limited to very specific landowners and developers. This approach would, however, be a first step towards winning back private investors to make any kind of long-term commitment.
Berlin is currently doing much that is possible within the legal framework to regain political influence on the housing market. However, its legal and financial leeway is limited at local or state level alone. An active long-term housing policy requires changes in federal tax legislation, supported by innovative land-use policies and urban planning. Neither the current capital nor subject subsidies have proven useful in implementing long-term social housing as they are repeatedly confronted with the logic of for-profit providers. Vienna has proven that affordable housing can generate income and achieve social goals, but it takes time for these investments to bear fruit, so temporary measures are neither sustainable nor effective in the long run.
When considering the meaning and function of interventions, the example of Vienna shows that it is not a question of state or market. Rather, it shows that the state can create framework conditions and incentive structures that make it possible to attract private actors to achieve social goals. Especially in long-term oriented housing construction, skilful state intervention can contribute to a better functioning of the overall market. However, this requires a long-term perspective and a political commitment to housing policy as a means to promote social policy goals rather than to correct temporary market imbalances.

DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.
ABOUT THE AUTHORS Susanne Marquardt is a guest researcher at the WZB Berlin Social Science Center. She is particularly interested in Welfare state research with a special focus on housing policies and inequality. She is currently pursuing her PhD on health policy and housing policy and how they reflect state capacity and interest group power.
Daniel Glaser is employed at the Department for Housing Promotion and Arbitration Board for Legal Housing Matters (Administration of City of Vienna). He is an architect, housing expert and interface coordinator for budget and controlling matters Furthermore he is responsible for international relations and research projects.