The resurgence of public housing provision in China: the Chongqing programme

Abstract Under Chinese neoliberalisation, restructuring of interactions between the market, the state and local governments has encouraged the latter to suppress public housing provision, resulting in serious housing problems for Chinese low-to-middle income households. Since 2007, the state has pursued a public housing revival, laying out various responsibilities for local governments to develop public housing. However, until recently, confronted with significant shortfalls in fiscal and land inputs, local authorities were largely unable to activate construction on an adequate scale. Nonetheless, between 2011 and 2013, the city of Chongqing applied a mode to supply public rental housing on a massive scale. This paper examines the administrative structure and reforms that have ensured the execution of the Chongqing programme. Findings show that specific political and economic incentives of the local officials have played important roles for realising the programme. Moreover, the municipality’s control of land supply and the market have also enabled an efficient cooperation between governmental and market actors for public housing provision.


Introduction
In recent decades, private real estate development has been prioritised by the Chinese government as a means of achieving unprecedented high-speed economic growth (State Council of China, 2003). Meanwhile, Chinese neoliberalisation has restructured interactions among the state, market and local governments, which has allowed local governments to withdraw land supply and finance for developing public housing, even though this tenure had been a mainstay for most households under the previous economic regime (Lee & Zhu, 2006;Wu, 2001;Yao et al., 2014;Ye & Wu, 2008). This housing provision bias has created a number of serious problems: shortages in low-cost (public) housing, income-based housing inequalities, poor housing conditions, etc. (Huang & Jiang 2009;Lee & Zhu 2006;Logan et al., 2010;Wang et al., 2012). Consequently, the state has been under severe pressures, and ostensibly in recent years, determined to focus on the resurgence of public housing, path differentiation within each society (Brenner et al., 2010). In most capitalist countries, weakening governmental intervention and more hybridised public and private sectors have become common features of housing provision (Mullins et al., 2014;Musterd, 2014;Ronald & Kyung, 2013). In China, neoliberalisation has been considered an incomplete process (Harvey, 2005), or 'neo-liberalization with Chinese Characteristics' (Wu, 2010), as it has featured both the expansion of market involvement and the maintenance of government regulatory controls. As figure one shows, on the one hand, the restructuring of statemarket relationships has brought about marketisation of housing provision and the intensified commodification of dwellings (Wang et al., 2012). On the other, the reorganisation of central-local government relationships has created a 'central-decentralized' housing provision structure in which the state strengthens its regulatory power and increases its share of revenue, while local governments organise actual policy implementations and even transfer hitherto public responsibilities for housing to the private sector (Chen & Gao, 1993). Therefore, in China, rather than a pure dergulation and market freedom shaping housing reform, expanding the role of the market has been directed by the central government, while local governments have continued to consolidate regulatory power in housing provision through intervening in the market, controlling urban planning and land supply, redistributing fiscal revenue, etc. (Fu & Lin, 2013;Zhang, 2002). The above restructured interaction among the market and the state in Chinese neoliberalisation has made local governments key intermediary agents linking the state and market in housing provision (Zhang, 2002).
As figure one demonstrates (Figure 1), the restructured interaction has also created a negative and complex context which has reduced local government incentives and capacities for expanding public housing provision (Chen & Gao, 1993). First, in context of the focus on developing the real estate market and promoting homeownership, neither local governments nor market actors have been interested in developing PRH (Huang, 2012). Moreover, both central and local government have withdrawn or minimised housing subsidies, while they have relied on market actors to invest in and manage public housing projects. Second, in terms of changes in housing provision environments, the process of decentralisation has nurtured reluctance within local governments with regard to the promotion of public housing. The uneven distribution of revenue and responsibilities have reduced local government capacities in distributing land and funds to public housing projects, while their empowerment has encouraged the redistribution of resources to promote urban industry (Zou, 2014).
From the above perspective, Chongqing has become a significant local variant in the ongoing interaction between urban neoliberalisation and socio-political objectives in China. It provides an important case for re-examining local government incentives and ways the market and government have interacted in developing social welfare, especially PRH. In the following section, building on the framework above, we more specifically examine how neoliberal transformations have changed both mechanisms and the environment of public housing provision, making both the government and market reluctant in developing public housing.

Changes in the provision structure
Under China's planned economy, housing provision was controlled by the single supplier, the state, under a communist ideology that asserted that social inequalities and class exploitation should be eliminated through the direct and centralised redistribution of housing, while the free market should be suppressed. As local implementers relied on the national budget support to construct public housing which was distributed to residents at very low rent (lower than its actual cost), the mode failed to provide public housing at an adequate scale and soon overburdened the state with substantial investment costs (Zhang, 2002).
During the Chinese economic transition, neoliberal reforms have thus brought various market actors together -such as employers, property developers, banks and private lenderswith local governments in the provision of public housing (Wang et al., 2012;Wu, 2008). In this integration, the state minimised intervention in policy formulation and strategic management. It no longer produced housing directly and only took the responsibility for housing the lowest income households. Since then, housing provision in China has become dominated by a three-layer provision mode which introduces private equity into public housing provision.
In this mode, the state stands at the top of a hierarchy and determines guidelines, affecting housing provision in three ways: land and housing planning, regulation on local government and controls on national banks. Local governments stand in the middle, establishing the institutional environment and strategies for housing supply, as well as negotiating with other actors in realising the construction and distribution of housing. Although they usually do not build houses themselves, local authorities play a core role in housing provision. They have controls of land distribution, taxation, bank loan guarantees and fiscal subsidies. Enterprises, mostly state-owned enterprises and property developers, play the role of implementers, raising funds from the capital market and even providing housing land. However, shifts in local government incentives (decentralisation), along with the ascendance of market actors and interests (neoliberalisation) have created a number of gaps in the provision of public housing (Li & Zhou, 2005). From 1998 to 2005, the annual national investment in all public housing only increased by 0.2 per cent while the national average house price nearly doubled (National Bureau of Statistics of China, 1998China, -2012.

Changes in funding public housing
The transformation of public housing finance started with privatising public housing, raising public housing rents, increasing tenants' salary levels, establishing housing accumulation funds, etc. Later, more approaches were introduced to harness more dynamic finance sources for public housing provision, moving away from revenue subsidies to interest and taxation reduction, housing allowance increases and opening up a diverse range of financial instruments. From 1998 to 2007, the share of public housing investment in the total housing investment decreased from 51.7 to 15.2 per cent (National Bureau of Statistics of China, 1998China, -2012. In the reformed funding structure, since 2003, the state provides 8-10 per cent of the total investment. It mainly outlines preferential banking policies and precise indexes on fiscal budget allocation and land distribution to ensure efficient policy implementations at the local level. Local governments are expected to organise the rest of the budget with the market. However, in reality, the main input from local governments is usually offset by a list of preferential policies, including policies of free land supply and cut tax, while the real cash input is very little. The market is expected to be the crucial source for funding public housing provision. However, unlike in other countries in which governments guarantee bank loans and ensure the repayment for public housing construction (Whitehead & Scanlon, 2007), housing developers in China do not gain sufficient and reliable guarantees from local governments. Instead, they mostly get bank loans through mortgaging their own assets, like their stock of land and housing, and their grants given by the government. Therefore, they are also reluctant in regard to getting their assets stuck in low-profit projects. Overall, few actors are willing to be active in public housing provision.

Changes in housing tenure structure
Housing tenure reforms in China were triggered by the Asian financial crisis of 1997, and required local governments to stimulate housing consumption for sustaining economic growth. Since then, housing has been treated as both commodity and welfare good in China. The state terminated the simplex distribution of public housing in 1998, and formed a more complex three-type public housing provision structure: featuring economic affordable housing (since 1994), capped price housing (since 2006) and low-rent housing (since 1998) ( Table 1). The former two are subsidised owner-occupied housing, making up over 60 per cent of the total public housing (State Council of China, 1998).
The tenure structure changes in public housing had made the housing provision uneven leading to, critically, a provision 'gap' and increasing levels of welfare exclusion. The majority of the low-to-middle-income households, people with salaries neither high enough to purchase subsidised owner-occupied housing nor low enough to apply for low-rent housing, have been excluded by the public housing system. The term 'sandwich population' , thus emerged as a result of the mismatch between the housing provision structure and the income distribution (Ying et al., 2013). Since 2007, the Chinese government has proclaimed its intention to correct the provision mismatch (State Council of China, 2007), with the arrival of the global financial crisis in 2008 being an important driver of the change. PRH has subsequently been designated as a more universal good to be provided at affordable prices, targeting both the sandwich population and residents qualified for other types of public housing. It is available to rent and for sale to sitting tenants. However, comparing with the other two types of public housing, as the target population of PRH is larger than the low-rent housing and the sale of PRH takes longer than subsidised owneroccupied housing, local governments have not been active in promoting PRH. Meanwhile, market sources are 'theoretically' available but 'practically' unavailable for PRH provision.

Environment for housing provision
Apart from transformations in the housing system, neoliberalisation in China has also brought about the division in incentives and power, undermining potential revenue and land supply for constructing public housing (Fu & Lin, 2013).

Imbalances in revenues and responsibilities
As figure one shows, the logic of decentralisation has been to transfer the development burden from the state to local governments while creating new incentives at the micro level in order to increase efficiency and production (Mok & Wu, 2013). Therefore, local governments have been allowed to gain a whole array of administrative powers over planning, public works maintenance and commercial administration (Lin & Zhang, 2015). However, the share of revenue and responsibility is not even between national and local government (Han & Kung, 2015).
Half of the revenue coming from locally controlled resources and locally generated fees and charges goes to the budget of the state, while smaller expenditures (around 15 per cent) are transferred back to the local level (Lin & Zhang, 2015). Between 1994 and 2011, the local fiscal expenditure has exceeded fiscal revenue, and the fiscal deficit has increased 29 times, from 1727 billion yuan to 50 729 billion yuan (National Bureau of Statistics of China, 1998China, -2012. Therefore, the decentralisation of responsibilities and recentralisation of the collection of tax revenue have overburdened local governments in terms of urban development and PRH provision. In 2011, the state planned to develop 10 million new public housing units, which required an investment of 1.3 trillion yuan, while it had only budgeted 0.103 trillion yuan in total (The Central Government of China, 2011). This represented a significant financing gap that had to be met by the private sector and/or local governments.
In terms of incentives, how state officials evaluate local officials has also had implications for the development of public housing. Criteria for promotion and dismissal has mainly been urban growth and market profits, rather than housing welfare improvements (Li & Zhou, 2005). This has undermined the administrative control of local governments' implementation of public housing policies, with local government officials often inclined to ignore the state requirements on public housing provision.

A growth-oriented land supply strategy
It has been argued that inter-governmental fragmentation has resulted in local governments having too much autonomy in deciding their land contributions for urban development, resulting a land shortage for developing public housing (Cao et al., 2008;Han & Kung, 2015).
With a limited national fiscal support, high pressures from the state and severe competition among cities, local governments have had to increase their revenue by all possible means. The opening of urban land market since the 1990s economic transition created a great opportunity for local governments to raise money through leasing land user rights to private developers. The income from commercialising land (or Tudi churangjin) belongs to the extra-budgetary revenue which goes primarily to the local budget, but with some flowing to the state. The process of generating revenue from land is termed as 'land finance' , a strategy whereby the government uses the anticipated price appreciation and the future added tax revenue of land earmarked for urban development to raise capital (Cao et al., 2008). In operation, once governmental planners reallocate the allowed use of a piece of land for development, the market value of the property rises sharply, dozens of times higher than the compensation, while the cost of land acquisition remains low as it is based the current use of the land. Thus, land finance is favoured by local governments as a major revenue source which can be used to further other political objectives. Between 1999 and 2011, the proportion of land finance in the total Chinese local revenue rose from 9 to 60 per cent (Ye & Wang, 2013).
Therefore, as the sole supplier of urban land, in the struggle to achieve a high GDP growth, local governments have been motivated to redirect land into infrastructure development and local industry, which not only boosts local GDP but also creates consistent income through taxes, and have refused to 'waste' land on housing development in many cases (Cao et al., 2008;Yao et al., 2014). The other reason for the tight supply of housing land is that the revenue income from leasing land to real estate developers is one-off, and local governments have to ensure an under supply of land for housing development in order to push land prices up.

Policy implementations at the local level
Above, we discussed how neoliberalisation in housing provision has suppressed local government incentives in developing public housing. However, as local governments are still subject to the state supervision to some extent, they have to 'fulfil' assigned public housing quotas. Therefore, below we examine various 'perfunctory' practices and a new mode of provision emerging across Chinese cities in the last decade.

Various 'perfunctory' practices
To ensure efficient local implementations, the state has outlined a series of policies for expanding PRH provision. For instance, in 2008, the state announced a '900 billion yuan plan' for developing public housing (State Council of China, 2008), and, again, in 2010, required that more than 70 per cent of new housing land should be supplied for public housing projects (State Council of China, 2010). In response to the state's 'unfunded' mandates, local governments have often overtly agreed but covertly resisted the national requirements. They have tried various 'perfunctory' ways to exaggerate the amount of PRH, while keeping their fiscal subsidies minimal.
In practice, the most convenient strategy has been to count other types of housing as PRH (see Table 1). For instance, some cities include low-rent housing, economic affordable housing and capped price housing as PRH (e.g. Guangdong, Henan and Fujian). Also, replacement housing provided for compensating households who lose their home to urban redevelopment (e.g. in Liaoning and Fujian) and employer-provided housing (e.g. Shanghai and Guangzhou) are often counted as PRH. Even some illegal housing built by the private sector on suburban-rural land or in urban villages, are sometimes counted (e.g. Shanghai, Shenzhen). As a consequence, the de jure scale of PRH supply has been larger, while the de facto supply has been very limited (Huang, 2012;Wang & Murie, 2011).
Another strategy is to require property developers to supply 5-10 per cent of their commercial property as PRH. This often works in cities with a sustained real estate market boom and is set as a precondition for property developers to acquire land for their private projects (e.g. in Beijing and Hangzhou). To encourage private sector investment in PRH programmes, local governments also promote a 'build and own' practice, which allows developers to hold on to the PRH they build as well as sell it off cheaply to eligible residents after 10-20 years (e.g. Beijing and Jiangxi). However, this strategy provides a very limited amount of PRH compared to the private real estate development it stimulates. Occasionally, local governments may also buy up unpopular private housing and old vacant public housing, and transfer it to PRH use, but again, the contribution is small.

A new provision approach among cities
The above 'perfunctory' practices have released pressures on local governments, but did not resolve fundamental PRH shortages. Consequently, in several cities, local governments started to experiment with a particular new approach in which special state-owned enterprises, namely FCEs, are embedded in a three-layer housing provision model for the financing and construction of PRH ( Figure 2). Essentially, the new mode is built for  restrengthening government involvement in the market to facilitate the construction of PRH. Thus, some local governments have realigned themselves behind public housing provision by establishing FCEs as a primary means for achieving state stipulated objectives, and their incentives and stock of land have become key elements for enabling the implementation of this mode.
In terms of the investment structure, in general, the state, local governments and FCEs contribute 10, 20 and 70 per cent of the total investment, respectively (Feng, 2011). FCEs receive local government subsidies and carry out contracts for public projects. As well as borrowing from the national bank, FCEs are authorised to issue government bonds, establish trust funds and gather loans from other social funds under the supervision of local governments. FCEs are also allowed to build profitable properties alongside PRH projects. In return, FCEs spend their profits in funding PRH construction.
FCEs work as intermediate actors between local governments and the private sector, taking responsibilities of supplying land for, investing in or organising PRH construction. They are authorised to simultaneously play the roles of land suppliers, investors, constructors or managers in PRH provision. In this way, the intermediate expenditure of the PRH construction can be minimised. The FCE works as a government body but is economically independent, differing from Work Units in the planned economy, which completely relied on government fiscal input to construct public housing. Therefore, in the new mode, the market offers an environment for FCEs to earn profits, while the government determines how to spend them. For local governments then, FCEs are much cheaper to run as they generate their own revenues (Feng, 2011).
Based on the new mode, cities establish their FCEs in their own ways. In cities with a large stock of land, FCEs are established from the state-owned land storage and supply enterprises, and thus FCEs can raise funds through leasing and mortgaging their land stock. In some big cities, like Beijing, FCEs are not authorised to store land. Local governments then provide land and other subsides to improve the asset base of their FCEs, with the later largely drawn from 'land finance' . Thus, the storage of land, whether the government or FCEs, and the power of FCEs have become fundamental to the success of the new mode. Below we examine how the municipality of Chongqing has specifically strengthened its government involvement in the market via FCEs, and has made specific adjustments to land supply practices in order to realise the Chongqing programme on a remarkable scale.

The PRH programme in Chongqing
The Chongqing programme targeted the construction of 40 million square metres of PRH, about 670 000 units (approximately 60 m 2 /unit), between 2010 and 2013. It has two FCEs, the company of Urban Construction Investment (Chengtoujituan) (UCI) and the Real Estate Company (Dichanjituan) (REC). Both FCEs are state-owned land storage and supply enterprises, and have been required to use their own assets to raise funds for the programme. The national fiscal input was far too small to support the whole programme, which had an initial estimated cost of 200 billion yuan. The Chongqing municipality worked to solve the remaining funding shortage through a series of measures. On the one hand, half the budget was met by granting preferential policies -on land and taxation -that favoured FCEs. The other half of the budget was met through a combination of a 20 per cent contribution from the municipality and a 10 per cent contribution from the state, with the remaining 70 per cent covered by bank loans, housing accumulation funds and social security funds (Feng, 2011).
To maximise FCEs' financing capacity, the PRH programme was run in a rather radical way. The municipality put forward the idea of the 'third hand': an approach combining public ownership with a market-oriented economy for satisfying public needs (Huang, 2011). Moreover, the municipality even mandated mortgage banks, the Housing Accumulating Fund and the Municipal Bureau of Finance to give priority to FCE's loan applications. Guaranteed by the municipality, FCEs were empowered to first use their stock of land as collateral for collecting bank loans, and then to repay loans after the construction was finished. The banks even reduced their interest rates as the municipality was not only the main client of the banks but also the administrator of the banks' other services.
According to Guo, the director of the PRH management bureau in Chongqing, the municipality outlined three approaches to repay the outstanding bank loans in the shortterm (Feng, 2011). First of all, an extra 10 per cent of land was distributed to FCEs for developing commercial properties, like banks, kindergartens and supermarkets, around PRH units. The rent derived from PRH tenants and commercial properties renters was estimated to be 6.6 billion yuan per year which would be sufficient to cover the annual interest of the bank loans (5 billion yuan) and management expenses of the PRH programme. Second, the sale of commercial properties was estimated to bring in 40-60 billion yuan, equivalent to 60-85 per cent of the bank loans. Third, as PRH can be purchased by sitting tenants after a five-year rental period, earnings from sales would also, in the short-to-medium term, be used to repay the loans.
The realisation of the new mode in the Chongqing programme was arguably due to two factors, the unusual incentives and powerful capacities of the municipal officials behind the FCEs and the large land stock of the FCEs. Below, we examine these two factors at length.

Unusual incentives and capacities of the municipal officials
As we have discussed earlier, in general, there is a persistent mismatch between the state's desire and local government needs in the sphere of public welfare (Faguet, 2004;Lee, 2000;Wu & Webster, 2010). However, in the Chongqing case, local and national governments ostensibly appear to have reached a successful agreement on the PRH provision. The agreement was related to the unusual incentives and power of the municipal officials.
To begin with, Chongqing is different from other cities. As the only Western Chinese city directly controlled by the state, Chongqing has been positioned as the major economic growth point of the vast, previously untapped, western China region, (Ministry of Housing and Urban-rural Development in China, 2010). Chongqing officials, therefore, have been under greater pressure, but have also been given more autonomy to make reforms (Lafarguette, 2011).
More specifically, the municipal officials in Chongqing, especially Mr. Bo Xilai, the secretary of the Communist Party's Chongqing branch between 2007 and 2012 (Cheng, 2013), were comparatively more powerful and particularly motivated in supervising FCEs in the construction of PRH. The PRH programme actually served Bo's political incentives rather than the public good. Bo is a former Chinese politician. Through his tenures as the mayor of Dalian, he had been particularly successful in establishing his political credentials at the 17th party congress in 2007, and between 2007 and 2012, he had planned to use his mandate in Chongqing as a platform to significantly further his own political ambitions, using his successes in Chongqing to bid for a seat at the top table of Chinese politics at the 18th party congress in 2012 (Gore, 2012). However, in 2013, he was subsequently convicted of corruption and sentenced to life in prison.
Bo established four core policy areas in his campaign to win local support (in Chongqing) as well as muster wider political recognition: improving local welfare conditions; maintaining a double-digit GDP growth rates; fighting organised crime and reviving the 'red culture' of Mao. In context of these objectives, the giant PRH programme became a crucial project that officials could tout in terms of promoting welfare equality and expanding welfare provision. In fact, the programme and reforms related to PRH also met Bo's economic goals by stimulating local economic growth and sustaining the necessary labour force for the continued growth of urban industry. During a rapid phase of urbanisation and industrialisation that has made it a base of manufacturing in China, Chongqing has had an exceptional need for workers and land compared to most other cities (Huang, 2010). The municipality estimated that investment in the PRH programme (110 billion yuan) would boost GDP by 400 billion yuan, and reforms related to the programme would also attract 910 000 untapped rural labourers to urban industry (Xinhua net, 2011), strengthening Chongqing's competitiveness in attracting large foreign investment and helping further its case as the centre of future manufacturing in China.

A distinctively large stock of land for raising funds
While most Chinese cities face land shortages when developing commercially profitable industries (never mind public housing development), rurally isolated Chongqing has an unusually ample stock of land, which it has increasingly begun to 'bank' . The core idea of land banking is to transfer the ownership of land to local governments so that they can make profits through directly leasing land to property developers or raising mortgages on the stock of land (land-based mortgage). As the largest and most economically underdeveloped city region in China (80 000 km 2 ), Chongqing has access to a large rural hinterland and a large surplus of rural workers. Thus, the unique combination of rural land and potential labour force provided significant opportunities for the municipality to accumulate swathes of undeveloped land at very low prices, as well as urbanise and industrialise rapidly (Lafarguette, 2011).
In Chongqing, UCI and REC act as FCEs, and are specific agents (established in 2002 and 2008, respectively) for collecting and consolidating plots of land. In practise, the municipality authorise FCEs to collect land located in planned construction zones and confiscate the unencumbered land supplement. By consolidating the land and making it 'developable' -by combining plots of land together and providing basic infrastructure (e.g. road access) -local authorities can effectively manipulate significant price increases. Consequently, by 2012, UCI and REC had already stored 200 km 2 of land, enough to meet expected Chongqing demand until 2022 (Feng, 2011). Guo Tangyong, the head of the PRH management bureau, indicated that the local government could comfortably dedicate 10 per cent of this stock of land for PRH construction (Feng, 2011). Moreover, if needed, the whole stock of land could be used for raising loans for the PRH programme.
Land banking is regarded as a costless strategy because the mortgage raised on the stored land is always used for buying more plots of land, and both the anticipated price appreciation and the future added tax revenue of the 'new-bought-in' land can be used to repay the loans of the 'previous-bought-in' land. Since the start of land banking in Chongqing in 2002, the land finance of the city rose from 0.2 billion yuan to nearly 90 billion yuan by 2012 (Time weekly, 2013). In order to promote land banking, some specific local reforms like the trade of 'dipiao' and hukou reform have been particularly important, as set out below.

The trade of 'dipiao'
In order to get more rural land stored in the 'bank' , in late 2008, Chongqing made a first, tentative attempt to link the rural and urban land markets via a warrant-like certificate, the 'dipiao' (Time weekly, 2013). The 'dipiao' is like a 'right to develop' , which allows property developers who want to build on greenfield sites close to urban areas to buy the 'right' from rural communities who have converted their construction land back to agricultural land elsewhere. The municipality promoted the trade of 'dipiao' as a means to balance urbanrural construction land usage without causing any loss of farmland (Romain, 2011). To some extent, it seems that all sides won in the trade of 'dipiao' . Property developers such as manufacturing firms, investment companies and commercial real estate companies, are buyers of 'dipiao' as it allows them to develop land in addition to their land quotas. Farmers, especially those whose houses are located in remote and poor areas and are keen on the scheme as they can cash in the value created by urban expansion, and receive around 85 per cent of the proceeds from the sale of 'dipiao' . Rural districts officials who represent farmers in land trades receive the remaining 15 per cent. By 2015, in Chongqing, the trade of 'dipiao' had provided rights to develop over 100 km 2 land (People, 2015).

Hukou reform and opening the PRH access to migrants
The other policy supporting land banking is the hukou reform. Hukou is the household registration system that was enacted in China in 1958, and which stated that for all PRC nationals in mainland China, their personal hukou status would be classified in terms of their original residential location ('agricultural hukou'/'non-agricultural hukou'). Restrictions placed on access to local welfare services by the hukou system have, for a long period, represented the biggest obstacle to migrants in establishing themselves in cities and been a major source of social inequality.
Following national hukou reforms, in 2010, migrants from rural parts of Chongqing have been encouraged to transfer their hukou from agricultural to non-agricultural status (Municipal People's Government of Chongqing, 2010). This involves migrants giving up their 'rural assets' , including their rights to housing and farmland in their official place of origin, and transferring them to the local authority. Three million rural residents are expected to have transferred their hukou status between 2011 and 2013, and 10 million by 2021 (People's Daily, 2011). This also meant that around 340 km 2 of rural land could be collected by the Chongqing municipality for potential construction and future land finance (Huang, 2010). In this regard, the reform actually worked to support the programme of land banking, as well as sustaining the labour force supply to the city. Moreover, through opening up PRH access to migrants, the programme has stimulated migrants to transfer their hukou status and settle down in the city centre. In the first year of the hukou reform, 40 per cent of the PRH were allocated to participants in the hukou reform scheme (China News, 2011).

Discussion
The above analysis demonstrates that conditions for the realisation of PRH programme in Chongqing were politically, economically and even geographically special. It seems that the Chongqing programme has ostensibly met national requirements for establishing public housing and stimulating the economy, while also improving the quality of people's living conditions (State Council of China, 2008). However, after Bo's downfall, many of the so-called 'innovative' reforms which facilitated the PRH programme have been proved to be fraud. Concerns should also be raised over equity issues and the longer term feasibility of these special reforms. Meanwhile the role of the municipality in restructuring its interactions between the market and the state should be critically discussed in order to better understand the influence of neoliberalisation in Chinese public housing provision.

Concerns over the land-based financing strategy
Since 1992, Chinese local authorities have started to outline legislation for storing, as well as leasing and mortgaging land for urban development. For decades, land-based finance had been treated as an efficient and convenient strategy to sustainable funding for local urban development (Ye & Wang, 2013). Meanwhile, scholars have also criticised the equity issues and the feasibility of these approaches (Ye & Wang, 2013).
First, according to regulation, land banking should be carried out on behalf of public interests. However the public, per se, are largely absent in this process. Regulations on land-based financing strategies prohibit individuals from directly transferring land user rights to property developers through the market. Moreover, in the process of land banking, farmers have had no right to dispose their plots but could be forced to give up their land and move to cities or other rural areas. Local governments determine lease prices and compensations in land banking and land leasing. They pay the compensation based mainly on the original use of land, which is much lower in value, while leasing land at prices based on the anticipated appreciation. Thus, earnings from land banking/finance are artificially exaggerated in an quasi-market environment. As landless rural migrants have gradually become aware of the increasing value of land in urban expansion, they have increasingly demanded equal citizenship to local residents and greater compensation from the government. Conflicts between land bankers and owners of land have deteriorated, and this may potentially undermine the excess profits from land finance.
Second, to keep the anticipated land value appreciation high, local governments have to maintain a stable market and positive economic growth. Once the market is disrupted or the economy deteriorates, fluctuations are also likely to affect the ability to raise funds through the land-based funding strategy. For example, during the second Asian economic crisis, land finance in Chongqing only increased by 11 per cent billion (from 28 billion in 2007 to 31 billion in 2008), but jumped to 131 billion by 2011 as the government increased its domestic investment to sustain the market (Chongqing Municipal Bureau of Statistics, 2008. Lastly, as cities keep expanding in China, big cities are facing growing land shortages for developing commercially profitable industries (never mind public housing development).
Problems of relying on land-based finance to fund urban development are thus likely to become a more serious concern (Cao et al., 2008).

Concerns over the trade of 'dipiao'
According to land management law, rural land is protected from being sold for developing urban industry, which makes urban land more valuable during rapid urbanisation, while the vast stock of rural land remains cheap and underused. Cities have recognised the value of rural land for urbanisation, but before using it, they have to legally transfer the status of land from rural to urban. Within this context, in 2008, Chongqing and the city of Chengdu, south-east of Chongqing, were first authorised as pilot regions to experiment with reforms to integrate rural and urban land markets. This gave Chongqing a chance to promote the trade of 'dipiao'. Later on, other local governments also started to trade 'dipiao' , but at a small scale, within their provincial boundaries.
However, there remains caution regarding the trade of 'dipiao' . First, if a quota target is set, the trade may potentially turn into a forced relocation programme, which might quickly be exploited by rapacious local governments and developers, while ignoring the will of land owners. Second, it is feared that farmers may be persuaded to sell their holdings at less than market rates unaware of land values. Third, as some landless farmers may not be able to adapt to the city life, the number of poor, landless peasants may soar, creating larger social instabilities. Another concern is that the quality of land brought back into cultivation in remote areas may not be as high as that lost.

Concerns over hukou reform and opening PRH access to migrants
In China, although, the requirement of including migrants in the social welfare system has been indicated in national documents, no concrete regulations have been enacted at the local level. The linkage between local hukou and local housing welfare systems is still largely intact (Huang, 2012). In Chongqing, the municipality promoted hukou transfer as a means to eliminate the division between rural and urban areas and to open access to full citizenship for migrants. However, in reality, the reform mainly targeted migrants who have already been working in urban areas for 1-5 years, which means the reform intentionally collected idle rural land without attracting more people.
Another concern is that reforms surrounding hukou may be storing up troubles for the future. For rural residents, their rural assets are passed on from generation to generation, meaning that the whole family can live on, and benefit from, the land in perpetuity. However, hukou reform takes their rural assets over permanently. Although migrants gain some one-off compensation and have more rights to public services in the city, most of them become economically insecure and vulnerable to periodic unemployment. In 2008, the global financial crisis provided some cautionary illustration of this for Chinese governments when around 20 million workers returned to the countryside (People, 2009).

Concerns over the political distortions
The unusual political incentives surrounding Bo Xilai's administration and the government-lead operation of FCEs have been key to the realisation of the Chongqing PRH programme (Cheng, 2013). They have also set the Chongqing programme apart from PRH programmes in other cities. After the downfall of Bo, many of his programmes were discontinued due to their association with the Chinese New Left (Gore, 2012). Although the PRH programme has been continued by the incumbent mayor, Mr Huang Qifan, it is not carried out in such an radical way anymore. Concerns have been raised regarding extreme or relatively unsustainable practices derived from the political distortions of the Bo Xilai era.
First, as goals behind the programme were mainly political and growth-oriented, since the downfall of Bo, the sustainability of PRH has been questioned and repayments to the high debt associated with the PRH programme has become a concern. In order to gain significant political traction before the 18th party congress in 2012, Bo changed the PRH project duration from 10 to 3 years (Official Website of the Chongqing PRH, 2010, 2011), meaning while many buildings were started, construction was not always finalised or finished off properly. With the downfall of Bo, and diminishing municipality authority, the construction of PRH also became much slower. By the end of 2016, only around 37 per cent of the project will have been finished and allocated to residents (Xinhua net, 2016). Moreover, at the beginning of the programme, the government announced that, from 2016, bank loans should have started to be repaid by selling PRH units to sitting tenants after a five-year renting period. However, by the end of 2015, no relevant regulation had yet been announced. Unless the government pursues other further financing strategies, it is unlikely that the loans will be repaid on time. Moreover, in context of an unfinished programme, FCEs might struggle to repay bank loans and acquire additional market capital. They may also have difficulties in getting additional loans to build further commercial and residential facilities in the newly built PRH communities. It has been revealed that most PRH neighbourhoods still have poor access to public transport, and lack kinder gardens, supermarkets, etc. A survey shows that satisfaction of PRH has been low (Kaifeng Foundation, 2011;Yan, 2014).
Another concern is whether such a programme has really benefited the general public. In 2011, at the first round of PRH allocation, of the 250 000 interested applicants, only 9 per cent were eligible. It was also reported that the rent of PRH units was too high for most low-income applicants to afford. Chongqing subsequently relaxed the rules to allow almost every income group to apply for PRH (Sina Finance, 2011). While this has made the PRH programme more popular, a large number of PRH has already been occupied by middleto higher income families. Moreover, in order to highlight the achievements of Bo, the municipality has often exaggerated official statistics in regard to the amount of PRH built. For instance, from 2010 to 2011, 18 000 of housing built by either non-public providers or remote counties was counted in the total amount of distributed PRH (110 thousand units) (Chongqing Morning Post, 2011). With the downfall of Bo, FCEs' ability to get financing decreases. Thus, the government has to rely more on the manufacturing companies to build more new PRH in order to maintain a stable production of PRH, and unfortunately, these houses will only be provided to employees of the factories and other commercial enterprises.

Concerns over FCEs and state-market interactions
In general, FCEs are legacies of both planned socialism and market-oriented capitalism. They are also comparable to the three primary actors of social/public housing provision in Europe: housing associations (e.g. the Netherlands), municipal housing companies (e.g. Sweden) and municipal departments (council housing in the UK but also Vienna). They work for local governments, like municipal housing companies and municipal departments, and also reinvest their profits on public affair related activities in a relatively non-profit way.
However, FCEs in Chongqing only appear to be economically independent like housing associations, but do not actually participate in market competition. Nor do they have to compete for banking finance. While it has ostensibly appeared as if the market voluntarily enlarged its participation in PRH, FCEs have used their authority to force market capital flows into the budget of the PRH programme. Banks and social funds are subjects of municipality authority and have to please government officials. Therefore, the function of FCEs in Chongqing still reflect many features of the former planned economy (Cheng, 2013). This kind of state-market interaction will be difficult to replicate in other cities unless there are also strong political and economic incentives. What's more, the function of the market is also not sustainable in such an interaction. This goes some way to explain why no other city has as yet managed to construct such a large amount of dwellings or repeat the apparent successes of the Chongqing programme.

Conclusion
To sum up, this paper has assessed ongoing changes in PRH provision in China, and in Chongqing in particular. We have examined the preconditions for, and provision mode of the PRH programme in Chongqing, which has been the first municipality to come close to achieving an adequate supply of PRH, as stipulated by national policy. Based on our analysis, three factors have been key to the realisation of Chongqing's PRH programme: the specific political conditions and incentives of municipal officials, the powerful capacities of FCEs and the distinctively large stock of land for raising funds for housing. However, there remain concerns over the feasibility and sustainability of this programme in Chongqing. These relate to land-based financing approaches; hukou and 'dipiao' reforms; radical policy practices driven by distorted political incentives under Bo Xilai, and the long-term enablement of socialist state-market interactions.
Positive publicity surrounding the Chongqing programme proliferated in China until the downfall of Bo. It has been regarded as an unprecedented success, and as such, the continuation of Chongqing's public housing programme has important implications for PRH development in China overall. First, it shows that, even under the current market-oriented housing system and the decentralised public housing provision scheme, there is potential for local governments to extensively expand their supply of public housing. Second, it illustrates how, in the Chinese context, the efficient development of public housing can be driven by the strong political and economic incentivisation of government officials. The state may have been sensitive to this, and since the emergence of the Chongqing programme, more policies have been announced that tie local government achievements in PRH to the evaluation of their performance. Consequently, there has been a wave of expansion in PRH provision across Chinese cities and provinces. However, there is also an awareness that the political and economic incentives behind the Chongqing PRH programme have been unique, and it is likely to be impossible for other cities to precisely mimic the Chongqing approach.
With respect to understanding of state-market interactions in providing public welfare, the Chongqing case indicates that, under the socialist regime, instead of extending the free market, local governments in China may choose to strengthen their control of the market to enable greater cooperation between governmental and market actors in reviving public housing provision. The reconfiguration of land and housing development around promoting urban labour supply, arguably reflects features of East Asian 'productivist welfare'(e.g. Holliday, 2000, Ronald andDoling, 2010). Chen et al. (2014) nonetheless, argues that the revival of public housing provision in China also signals transformation from a 'productivist welfare' to 'developed welfare' approach. The municipality of Chongqing has also been entrepreneurial (Duckett 2001), operating within the parameters of 'local state corporatism' or 'local state marketism' (Walder, 1995;Fu & Lin 2013;Oi, 1995). The analysis also meets Wu's argument (2010) that Chinese style neoliberalisation has again led to unpredictable outcomes in terms of the interaction of capital and the state in the sphere of housing, with public housing needs met by adaptations to the market within China's urban growth machine.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
Richard Ronald's contribution to this article has been supported by the European Research Council [ERC Starting Grant HOUWEL, Grant Agreement No. 283881].