1 Introduction

China’s rapid urbanization is mainly due to the migration of millions from the countryside since the economic reforms started in 1978. These rural-to-urban migrants (Nongmingong in Chinese) leave to work in urban areas as cheap manual laborers (Du et al. 2005). However, they rarely have access to the destination’s household registration system (hukou in Chinese) on which the social welfare system is based. With no prospect of social security in old age, they return to their place of origin when they can no longer do heavy labor. The institutional constraints inherent in the hukou system prevent rural migrants from staying permanently at their destination, which reduces their enthusiasm to purchase a dwelling there.

Starting in the late 1990s, however, the national government’s urbanization ambitions and the increased freedom to migrate within the country constituted a force for change (Wang 2010). These ambitions prompted efforts to improve China’s institutional system. A significant outcome is the relaxation of hukou constraints (Chan and Buckingham 2008). Some provinces no longer distinguish between an urban hukou and a rural hukou: no matter where people live in a municipality, their municipal-level hukou is the same (Chang 2010). Furthermore, if rural migrants meet stipulated criteria, they can exchange their hukou of the place of origin for a hukou of the destination place (Wang 2010). Moreover, eligibility for the social welfare systems is gradually being separated from one’s hukou status (Yu and Ding 2010). This trend is particularly evident in the urban social insurance system, which has been open to rural migrants since the early 2000s (Nielsen et al. 2005).

However, there are large differences among the destinations, especially in their policy on access to municipal-level hukou. That is because each municipality’s access policy reflects its particular stage of development as well as the advantages of a local hukou status in relation to the institutional reforms (Wang 2004). More-developed municipalities, where the hukou is highly valued because it offers residents a range of facilities and social programs, still put strict conditions on access, while less-developed municipalities offer easier access (Wu et al. 2010). There are also considerable differences among rural migrants with respect to their home hukou status. They may be divided into three types: intra-municipality; inter-municipality; and inter-province. These types of rural migrants are expected to respond differently to various reform policies.

This paper explores how the institutional changes emanating from the nationwide hukou reform affect home ownership among rural-to-urban migrants. Primarily, we expect to find that, by lowering the institutional threshold for home ownership among the migrants, this reform will make demographic variables more important in explaining the rate of home ownership. Furthermore, as divergence in the value of the welfare program among provinces and municipalities will presumably reflect differences in the level of local institutional access, we expect to find that the influence of each reform package will differ among the destination places, thereby creating differences in the rural migrants’ patterns of home ownership. To our knowledge, these two suppositions have not yet been subjected to empirical investigation after the hukou reform. That is precisely what we intend to do in the course of our research. The present study seeks to start bridging that gap in knowledge by posing the following research questions: What are the effects of institutions and demographic characteristics on rural migrants’ home ownership rates after the hukou reform in China? and Which differences occur between destination municipalities in the same province in this respect?

The paper starts with a review of the literature on rural–urban migration, hukou status and hukou reform, and rural migrants’ tenures in destination municipalities. The empirical part presents a regression analysis to model home ownership rates, using data from a 2009 survey on more than 3,000 rural migrants in Jiangsu province. The subsequent section summarizes the effects of institutions and demographics on home ownership rates among individual rural migrants and compares the divergence in its prevalence between municipalities with a different extent of hukou reform. The paper ends with an exploration of some directions for further hukou reform that might gradually eliminate the institutional barriers to rural–urban migration in China.

2 Literature review

2.1 Rural–urban migration and the hukou system before reform

A sizable body of literature has expressed concerns about China’s rural–urban migrants. Their migration is usually ascribed to a surplus of rural labor and an income disparity between the city and the countryside (Du et al. 2005). But China’s ongoing economic reforms have also led to serious regional disparities. In a geographic respect, there are three kinds of rural–urban migration (see Fig. 1): intra-municipal (relation 1: laborers from a rural area go to an urban area in the same municipality); inter-municipal (relation 2: rural laborers from less-developed municipalities go to urban areas in more-developed municipalities in the same province); and inter-provincial (relation 3: rural laborers from less-developed provinces go to urban areas in more-developed provinces).

Fig. 1
figure 1

Three kinds of rural–urban migration

Full size image

However, that population flow has been cushioned by a tight system of migration control. A household residency registration system (hukou, the term used for both the system and an individual’s status within it) was reinstated in the 1950s to prevent indigent peasants from flooding into China’s war-torn urban areas (Chan 2010a). At the very beginning, every person was registered at a specific place, which referred concomitantly to three geographic attributes: urban/rural; municipal; and provincial. Under this system, one’s hukou comprised three indicators: urban or rural (attribute one); which municipality he/she belonged to (attribute two); and which province he/she belonged to (attribute three). Newborns inherited their parents’ hukou status, no matter where the mother gave birth or where the family was actually living. Consider the example of a baby born legitimately in urban Nanjing to a migrant couple working there. Expressed in those three attributes, the parents had a Jiangsu (province-level)–Yangzhou (municipality-level) rural (rural or urban) hukou; the infant would inherit their Jiangsu–Yangzhou rural hukou and would not obtain a Jiangsu–Nanjing urban hukou.

The value of any particular hukou lies in the social welfare provisions that come with one’s hukou status. Provincial and municipal governments offer public services according to the digits denoting the local hukou (Wu et al. 2010). For example, the baby with the Jiangsu–Yangzhou hukou enjoys the provisions of Yangzhou’s social welfare system but not Nanjing’s, even though the infant was born and raised in Nanjing. The child has access to Jiangsu’s provincial benefits because a hukou for either of these municipalities falls under the Jiangsu hukou.

The reason for the inter-area difference in the field of social welfare is complicated. China’s municipal governments get most of their revenue from taxing local enterprises and selling urban land (Zhang 2009). In any given province, the more-developed municipalities can extract higher tax revenues from local enterprises. They can also get more money from the sale of urban land for the construction of market sector housing, due to the high price of land, which is closely connected to the high price of housing. Accordingly, the revenues of more-developed municipalities are higher than those of less-developed municipalities. Therefore, regarding the provision of public goods, the former can spend more on each citizen with a municipal-level hukou, can maintain a high standard of social welfare services, and can provide more public infrastructure than the latter (Wu et al. 2010). At the regional level, the provincial government’s revenue comes mainly from creaming off the municipal governments’ revenue (Zhang 2009); thus, rich provinces can offer better social welfare programs than poor provinces. Under these conditions, the hukou of more-developed places is obviously more valuable than that of less-developed places (Li 2010).

Regarding the difference between an urban and a rural hukou, China’s municipal structure has a two-track social welfare system connected to an individual’s hukou status. Someone with an urban hukou is entitled to public amenities and social services like health care, housing, pensions, and education for their children (Treiman 2012). But in rural areas, the rules are different, especially for social security. A rural hukou gives a peasant land-use rights, reflecting a thousand-year-old tradition of relying on agriculture (and grown children) for one’s security in old age (Chan 2010b). Therefore, in terms of the three attributes of a hukou status, one’s social welfare provisions depend on three differences: between urban and rural areas; between municipalities; and between provinces.

The main thrust of our analysis is to unravel how these welfare differences relate to rural–urban migration. This mechanism may be visualized as a disconnection between institutional space and the geographic space of an individual. Ideally, people should be able to enjoy the social welfare amenities of where they live; this presupposes a direct relationship between the geographic space and the institutional space (see relation 1 in Fig. 2). The interjection of hukou disrupts the interaction between those spaces (relation 2 in Fig. 2). Access to a social welfare system is based on one’s hukou status, and people are entitled to the social welfare amenities available where their hukou “lives.” After migration (relation 3 in Fig. 2), if a migrant’s attempts at changing the home hukou into that of the destination fail, he/she cannot enjoy the social welfare amenities of that destination. Few migrants manage to transfer their hukou due to the strict constraints on altering the location of one’s hukou status (relation 4). In that sense, the hukou system creates a disconnection between an individual’s geographic place of residence and his/her social welfare provisions (relation 5). This is what we call a mismatch problem. The analysis will place the empirical evidence we have gathered against the backdrop of this conceptual model that we have constructed from the literature. And the mechanism of mismatch set forth above can serve well as that model.

Fig. 2
figure 2

The mismatch problem during migration brought about by hukou. Solid two-way arrow line 1 Match between institutional space and geographic space. Solid two-way arrow line 2 Interjection of hukou. Solid one-way arrow line 3 Out migration. Solid one-way arrow line 6 Return migration. Dotted one-way arrow line 4 Institutional change (suppositional hukou transfer). Dotted two-way arrow line 5 Mismatch between institutional space and geographic space

Full size image

The dynamics of mismatch play out in the three kinds of rural–urban migration. Intra-municipal rural migrants can enjoy all municipal social welfare amenities at their destination. The mismatch they face is that in many places they still have limited access to urban facilities. Inter-municipal rural migrants can only enjoy provincial social welfare amenities at the destination, so they face two mismatch problems: one between municipalities and one between the urban and the rural area. Inter-provincial rural migrants hardly enjoy any social welfare amenities at the destination, and their mismatch problems are the most serious. As a result of the mismatch, and without the prospect of security in old age, a migrant has to engage in return migration in order to access the social welfare system at the place of origin. As a consequence of the mismatches brought about by the hukou system, the incidence of permanent rural–urban migration has been reduced. Moreover, given the intention to return to their place of origin (see relation 6 in Fig. 2), rural migrants tend to leave home without giving up their land (Cai 2003). In that sense, the hukou system may be said to have slowed the pace of urbanization in China (Au and Henderson 2006).

2.2 Hukou reform, alteration in hukou status, and urban insurance

The national government intends to stimulate those aspects of urbanization that fuel the domestic market in the hope of sustaining economic growth. Yet it was not only the foreseen economic impact that led to the hukou reform. The reform was also a response to the rising political appeal for the freedom to migrate within the country (Wang 2010). The institutional changes entailed the decentralization of fiscal and administrative powers. These have been moved down the hierarchy from central government to provincial authorities, then on to the municipalities since the late 1990s (Chan 2010a). Provincial and municipal governments also expect the better-off rural migrants to purchase urban housing, to invest in the local economy, and to settle permanently in the municipalities. The authorities encourage such local bonding for its assumed positive effects that help in the fierce competition over economic growth with other provinces and municipalities (Zhang 2009).

According to our conceptual model of the impacts of hukou on migration (sketched above in Fig. 2), eliminating the mismatch problem is the key ambition propelling the hukou reform. Theoretically, there are three main approaches to achieve that goal. The optimal way is to abolish hukou status, either completely or at least in one of its three attributes (urban/rural, municipal, provincial). The next best way is to open up more avenues to alter one’s hukou. The least appealing way would be to separate some of the social welfare provisions from the hukou status. In practice, in the course of the hukou reform, provincial and municipal governments have taken all three approaches to varying degrees, thereby restraining return migration.

The first approach has been applied in some provinces, including Jiangsu, whereby the distinction between rural and urban hukou has been abolished (Chang 2010). This measure has removed one of the three attributes of a hukou status, to the benefit of intra-municipality and inter-municipality migrants. It is no longer necessary for intra-municipality rural migrants to change their hukou status in order to take part in the entire social welfare system to the same extent as urban residents do. This measure has also weakened the role that farmland played in providing security for peasants in old age and thereby eased the resistance in the origin municipality to expropriation of farmland for urbanization. The problem of mismatch between urban and rural has been ended for inter-municipality rural migrants, who are now only confronted with a difference in the welfare provisions between municipalities. However, the effects on inter-province rural migrants are ambiguous, since not all provinces of origin have adopted the same policy.

Regarding the second approach, any change in one’s hukou must be approved by the authorities at both the place of origin and the destination. Before the reform, it was difficult to change one’s hukou; the main avenues were through obtaining a civil service job after finishing higher education or performing military service (Chan and Buckingham 2008). These avenues are beyond the reach of most rural migrants. Since the reform, three more channels have been open to inter-municipality and inter-province rural migrants to enter destination hukou systems. One is a housing qualification (applicant owns a home of a certain size in the destination area); the second is an insurance qualification (applicant has been enrolled in local urban schemes for a certain period); and the third is a marriage qualification (applicant is married to a resident with a local hukou) (Jiangsu Provincial Government 2003). Provided that an inter-municipality or inter-province rural migrant meets one of the qualifications, he/she can exchange a home municipality hukou for a destination municipality hukou. And with a municipal-level hukou at the destination, an inter-province rural migrant automatically gets the corresponding provincial-level hukou.

Compared to the other two qualifications, home ownership offers inter-municipality and inter-province rural migrants the best chance to change their hukou status. For instance, an insurance qualification has only been adopted in three of the eight main destination municipalities in Jiangsu province. Nor is the marriage qualification a straight shot: rural migrants do not easily find a marriage partner among local residents due to their low social status in urban China (Davin 2005). However, municipalities differ in the way they interpret the housing qualification, reflecting the difference in the value of any particular municipal-level hukou. It is relatively high in the more-developed municipalities; thus, newcomers are expected to invest more in the local economy in order to obtain a municipal-level hukou (see Li et al. 2010). Conversely, Wu et al. (2010) have argued that governments of less-developed municipalities are inclined to enact more lenient hukou policies in an effort to absorb cheap labor to benefit local enterprises.

Inter-municipality and inter-province rural migrants who meet the housing qualification then have to decide whether to go ahead and change their hukou status. They face trade-offs. Once inter-municipality and inter-province migrants shift their hukou from their home municipality to their destination municipality, they give up their farmland use rights. Doing so voluntarily, they do not receive the compensation they would expect whether the farmland had been expropriated (Lin and Ho 2005). Many migrants therefore wait until they can “sell” their home hukou at a good price before applying for a hukou at their destination. For example, in our 2012 survey in the municipality of Yangzhou, fewer than half of the migrant homeowners had applied for a Yangzhou hukou. The reluctance among rural migrants to give up their home hukou in turn threatens the success of the receiving municipalities’ efforts to ease hukou access.

The amount of money needed to purchase market sector housing puts home ownership out of reach for most laborers. On average, a market sector unit costs more than 20 times the average annual income in China’s municipalities (Wang 2012). Moreover, before the hukou reform, rural migrants without a local hukou could not get a mortgage (Wu 2006). If governments intend to sell more urban dwellings at such high prices, the hukou reform would have to open up access to mortgages. For instance, in 2007, Jiangsu province made an institutional improvement enshrined in guidelines for “Solving Rural–urban Migrants’ Difficulties” (Jiangsu Provincial Government 2007a). This milestone legislation permits certain inter-municipality rural migrants to participate in the housing provident fund. Although the criteria for eligibility are fairly strict, inter-municipality rural migrants now have access to sources to overcome the financial impediments to home ownership. However, inter-province rural migrants are still barred access to housing loans. The reason may be that a transprovincial mortgage requires cooperation between provinces and thus lies outside the jurisdiction of the provincial government. Inter-province rural migrants therefore rarely meet the housing qualification.

The third approach is to separate participation in an urban insurance scheme from one’s hukou status. If covered, a migrant can count on a secure livelihood in old age at the destination (Davies and Ramia 2008). For example, Jiangsu Provincial Government now allows rural migrants to purchase commercial insurance policies from state-owned insurance companies regardless of their hukou status (JSHRSS 2007). At the municipal level, there seems to be no direct link between the government’s budget and the migrants’ insurance; municipal authorities are rarely interested in developing an insurance policy of their own. Instead, they usually operate the scheme of the provincial government.

An additional stipulation addressing the social security concerns of intra-province migrants has been implemented to uphold the value of the provincial-level hukou. Since 2007, rural migrants with a Jiangsu hukou can transfer their insurance for free among places within the same province (Jiangsu Provincial Government 2007b). In contrast, if inter-province laborers migrate to another destination or go back home, they have to drop their existing coverage and purchase insurance at the other place. That clause deflates the enthusiasm of inter-province rural migrants to participate in an urban insurance scheme at their destination.

Summing up the problem of mismatch between social welfare and hukou, we expect to see differential effects of the hukou reform (see Table 1). The welfare of intra-province rural migrants—including intra-municipality and inter-municipality laborers with a provincial-level hukou at their destination—has improved. A tangible improvement encourages rural migrants to call that destination home. They are then more likely to purchase a market sector dwelling or save up for one. For the inter-province rural migrants in contrast, the effects of the reform are not so obvious. Their status prevents a permanent stay at the destination, so they will have to engage in return migration, as before. They are consequently inclined to invest in housing in their hometown, not in their current place of residence.

Table 1 Effects of hukou reform on mismatch problems for three types of rural migrants
Full size table

2.3 Determinants of rural migrants’ home ownership at destination municipalities

The determinants of rural migrants’ housing tenure at destination municipalities have been extensively discussed, mostly in light of empirical work on different municipalities before the hukou reform. Diverging from the emphasis on demographic or economic factors prevailing in research on Western countries (e.g. Deurloo et al. 1987), most research on China has emphasized the effects of institutional constraints.

A local urban hukou (see Huang and Clark 2002; Huang and Jiang 2009; Jiang 2006; Logan et al. 2009; Wu 2004), in addition to a higher-ranked occupation (see Li 2000), and a public sector job (see Zhou 2011) are associated with home ownership. Most residents with a local urban hukou could enter into home ownership at a low cost through two channels: they either inherited a property or obtained it at a low price by purchasing collective-owned housing such as work units (Logan et al. 2009). The former channel is largely closed to rural migrants, since there are few urban dwellings for them to inherit. For the latter channel, without having a local urban hukou, rural migrants could not be official employees of work units. As a consequence, the work units would not assign collective-owned housing to them (Huang 2003). Only those already living in collective-owned housing were given the option to purchase work units housing, thereby attaining home ownership (Logan et al. 2009). In short, if rural migrants enter into home ownership, they do this mainly by purchasing market sector housing.

The demographic effects have been elucidated by Huang and Clark (2002). Working with data from a nationwide survey conducted in 1994 among residents with a local urban hukou and migrants in 1994, they found that age (older), household size (big family), and a higher household income are associated with becoming a homeowner. Using data from Guangzhou for the period 1992–1994, Li (2000) found that only higher income was associated with home ownership; other life-cycle variables were statistically not significant. Later, Huang and Jiang (2009) examined home ownership in Beijing using data from 2000 and found that a big household was positively related to home ownership. Wu (2004) examined migrants’ housing tenure in his own survey of 2000 in Beijing and Shanghai. He found that age, education level, marital status, and household income were not statistically significant with respect to owning a dwelling.

In sum, all studies report that home ownership among rural migrants was mainly institution-dependent before the hukou reform. The numerous studies on life-cycle variables have led to considerable debate instead of consensus. Yet it is generally accepted that the effects of an individual’s life-cycle position are only moderately important, as the role of demographics has been cushioned by institutional constraints. To cope with the mismatch problem, most rural migrants have engaged in return migration and have become homeowners back in their hometown. Socio-demographics can predict when a person will become a homeowner but not where. In this sense, institutional constraints do not explain all the effects of demographics on home ownership, only those at a certain place—the destination municipality.

However, the hukou reform is expected to change that and we pursue two main hypotheses that express our expectations of the changes. Firstly, with the decrease in the institutional impediments, we posit that home ownership among rural migrants after the hukou reform will be linked to life-cycle factors as well as to attaining a local hukou and participation in welfare programs. Furthermore, we expect to find that in the same province, the more-developed municipalities have maintained rather strict hukou constraints after the reform. This may be formulated as our second hypothesis: in the less-developed municipalities, where access to social programs is relatively easy, the effects of institutional factors on home ownership among rural migrants are less important than in the more-developed municipalities of the same province.

3 Case study, dataset, and research design

3.1 Case study area

Jiangsu province is a highly suitable region for a case study to evaluate rural migrants’ home ownership after the hukou reform. The region includes a booming coastal economy that attracts large numbers of migrant workers from impoverished regions of China. Moreover, since the hukou reform began, the national government has used this province as a model to test its new policy of relaxing constraints on hukou access and rural residents’ participation in the urban insurance system (Fig. 3).

Fig. 3
figure 3

Location of the eight municipalities in Jiangsu province, China

Full size image

In terms of hukou access, the provincial government started the reform in January 2003 by setting guidelines for the hukou reform (Jiangsu Provincial Government 2003). As set forth above, in order to apply for a local hukou, a rural migrant has to hold a housing, insurance, or marriage qualification. Each municipal government was required to revise its hukou access policy based on these guidelines. In the meantime, eight have done so: Wuxi (April 28, 2003); Suzhou (April 30, 2003); Taizhou (June 30, 2003); Zhenjiang (July 1, 2003); Changzhou (July 1, 2003); Nantong (October 9, 2003); Yangzhou (June 1, 2004); and Nanjing (June 19, 2004). Table 2 compares the requirements posed by these eight municipalities in 2009 when our fieldwork was finished.

Table 2 Hukou reform policies of eight municipalities in Jiangsu: requirements to obtain a local hukou
Full size table

According to the guidelines of the Jiangsu Provincial Government, the three pathways are theoretically open to rural migrants. In practice, however, their success depends on how municipal governments apply these rules. It may be easier to obtain a housing qualification than the other two because all of the municipalities look favorably on rural migrants’ purchase of market sector dwellings as it means investment in the local economy. In addition, each destination municipality has its own reasons to emphasize or adapt particular details of these three qualifications. Presumably, the difference in their level of development goes a long way toward explaining the variation in the adjustments municipalities make in the eligibility criteria. To verify this supposition, we have ranked the municipalities’ developmental status by their GDP and GDP per capita, two indicators that are commonly used for China. In that respect, the most highly developed urban areas in Jiangsu province are Nanjing, Suzhou, and Wuxi, respectively holding the top three positions (Jiangsu Statistics Bureau 2010). Concomitantly, their threshold for hukou eligibility is relatively high. Besides requiring a higher amount of investment in home ownership, they do not offer the channel of insurance qualification. In contrast, the criteria set by other municipalities are less restrictive (cf. Table 2). They apply more lenient criteria to attract cheap labor, and the rural migrants will not have to spend as much on a dwelling to gain hukou access.

Rural migrants in Jiangsu province mainly participate in five kinds of urban insurance schemes: work injury, pension, medical, unemployment, and maternity insurance (Nielsen et al. 2005). In 2007, a comprehensive official supplement to the provincial government’s Guidelines was issued (Jiangsu Provincial Government 2007a). That document clarified the conditions under which rural migrants could participate in various insurance schemes.

This supplement divided rural migrants into two groups: those who signed official contracts with employers and those without official contracts, including the self-employed (JSHRSS 2007). Regarding the former, the employers were bound to several conditions: they should participate in work injury insurance on behalf of the rural migrants and pay the entire premium for their employees; they should notify the migrants of their rights to an urban pension and unemployment coverage and pay part of the premium; and in terms of urban medical insurance and maternity benefits, the employers should negotiate with rural migrants on the possibilities for participation. Rural migrants without official contracts participate in these insurance schemes voluntarily, as local workers do, and negotiate directly with the insurance companies on the details. Both the guidelines and its explanatory supplement are intended to regulate employers’ behavior such that the rights of rural migrants’ participation in social insurance schemes are guaranteed. And it is up to the rural migrants themselves whether or not to purchase urban insurance, with the exception of work injury insurance, which is paid for by the employer.

3.2 Dataset and research design

Our analysis is based on the 2009 survey that was administered by Nanjing Normal University to more than 3,000 rural migrants in eight municipalities in Jiangsu province: Nanjing (438), Suzhou (481), Wuxi (390), Changzhou (402), Zhenjiang (279), Yangzhou (277), Nantong (322), and Taizhou (217). After excluding invalid questionnaires, the sample size was 2,806 (250 individual homeowners and 2,556 non-owners).

The analysis was performed with binary logistic regression models, the dependent variable being home ownership among rural migrants (owning equals 1; not owning equals 0). The predictors include socioeconomic and demographic variables (age, gender, education level, household size, and personal monthly income) and institutional factors (urban pension insurance, medical insurance, unemployment insurance, work injury insurance, and province-level hukou). We excluded maternity insurance from the analysis performed on the whole sample. Our survey recorded the province-level hukou of the respondents, distinguishing between rural migrants with a Jiangsu hukou (intra-municipality rural migrants and inter-municipality rural migrants) from those with a non-Jiangsu hukou (inter-province rural migrants).

A descriptive overview of the variables in the dataset is shown in Table 3. Less than nine percent of the respondents proved to be homeowners. The average age for the entire sample is about 34 years, and more than 60 % of the respondents are males. An educational attainment of 9 years of schooling is compulsory in China. After that, students can choose to graduate and enter the labor force or to continue their education. Personal income has two categories: less or more than 1,600 Yuan per month. The average monthly income for urban employment in Jiangsu province was around 1,570 Yuan in 2009 (Jiangsu Statistics Bureau 2010). Household size refers to all family members living together in the destination municipalities, including migrant couples, children, elderly parents, and siblings.

Table 3 List of variables
Full size table

In the analysis, we first performed binary logistic regression on rural migrants’ home ownership for the eight municipalities together to demonstrate the effects of the independent variables, namely demographics and institutional factors. Then, we divided the eight municipalities into two groups based on their developmental status and institutional thresholds. Group 1 consists of the more-developed municipalities of Nanjing, Suzhou, and Wuxi, all having a high threshold for participation in the institutions. Group 2 consists of the less-developed municipalities of Changzhou, Yangzhou, Zhenjiang, Taizhou, and Nantong. We modeled the rural migrants’ home ownership for each group and compared the effects of the institutional factors. Finally, we applied regression models to each municipality separately, using the independent variables of two key institutional factors, namely a province-level hukou and urban pension insurance. This analysis showed the inter-municipality differences with respect to these institutional factors.

4 Regression analysis

4.1 Effects of institutional variables and demographics on rural migrants’ home ownership

Model 1 reveals the respective contribution of the institutional factors and socio-demographics for all eight municipalities together, resulting in a Nagelkerke’s R square of 0.299 with 11 degrees of freedom (Table 4). The results show that age, gender (female), higher educational level, large household size, higher personal income, Jiangsu hukou, pension insurance, and unemployment insurance are all significant predictors of home ownership among rural migrants. The effect of medical insurance is statistically insignificant, while work injury insurance has a negative correlation.

Table 4 Binary logistic regression models of rural migrants’ home ownership
Full size table

When the institutional effects are viewed in detail, it appears that having a Jiangsu hukou increases the likelihood of owning one’s home, which is consistent with findings of other studies. This effect is probably a result of two conditions. For one thing, compared to inter-province rural migrants, intra-province migrant workers are more strongly motivated to invest in housing in Jiangsu province, since it is their home province (De Brauw and Rozelle 2008). In addition, since the reform, a Jiangsu (but non-local) hukou allows rural migrants to participate in the housing provident fund or to obtain a mortgage loan. Either option would solve their financial problem of how to pay for home ownership. Under these conditions, rural migrants with a Jiangsu hukou are more likely to attain home ownership. In that sense, the hukou reform creates disparities between intra-province and inter-province migrants.

Participation in the urban pension insurance scheme is also a positive predictor. The odds of being homeowners are nearly three times as high for rural migrants who have pension insurance than for those who do not (see Table 4). It is likely that the guarantee of an old-age pension will clinch a migrant’s decision to buy a dwelling and settle down on a permanent basis. And in some municipalities—namely Changzhou, Nantong, and Taizhou—participation in an urban pension scheme for several years allows a rural migrant to apply for a local urban hukou (see Table 2). Similarly, having unemployment insurance, which serves as a guarantee of steady income, is also a positive predictor of home ownership (see Table 4).

In contrast, work injury insurance has a negative correlation with home ownership, possibly due to the type of employment of the insured migrants. People working in construction and manufacturing should take out injury insurance, mainly because of potential physical harm at the workplace (Nielsen et al. 2005). However, rather than buying a dwelling in the market sector, those workers usually live in sheds at the construction sites or in factory dormitories (Li et al. 2009). Medical insurance is statistically not a significant factor, probably because Jiangsu’s policy would usually only cover serious illness (not common ailments) (Jiangsu Provincial Government 2008). Therefore, it is not very attractive to rural migrants; the prospect of eligibility for medical coverage would not induce them to settle down and pursue a local hukou.

Regarding demographics, age and higher income have a positive correlation with home ownership. These variables are reflected in the savings accumulated over the course of one’s labor migration. Another variable with a clearly positive correlation is a higher level of education, which usually leads to a higher-ranking occupation and higher income. The number of family members living together has a very strong positive effect on home ownership, probably because market sector housing is usually more spacious than factory dormitories and low-rent accommodation. This finding is consistent with that in the review by Clark et al. (1994) regarding the life-cycle effects on tenures. They found that an increase in the number of family members would trigger residential mobility.

It is interesting to note the gender difference in the propensity of rural migrants to enter into home ownership. We found that being female has a positive correlation with home ownership. Given China’s strong tradition of the patri-family and marriage roles, husbands usually have a higher socioeconomic position than their wives (Davin 2005). Due to gender imbalances, female migrants are likely marriage partners for local males, and those males who own a house are more likely to be successful in their pursuit of matrimony. Conversely, male migrants are not likely marriage partners for local female residents.

It is instructive to compare our findings with those of other studies on home ownership, including some dealing with China’s rural migrants before the hukou reform. In terms of age, educational level, income, and household size, our findings resemble those drawn from research in market economies. We too found that individuals who are older, more highly educated, with a higher income and a big family are more inclined to own private housing (cf. Deurloo et al. 1987). One exception concerns the effect of gender, which reflects China’s marriage culture. Thus, in view of the analysis of Model 1, our first hypothesis appears plausible: After the hukou reform, rural migrants’ home ownership is determined by both life-cycle factors and participation in institutions.

4.2 Logistic model for inter-municipality difference in rural migrants’ home ownership rate

Our second hypothesis is focused on the disparate influences of the reform package on the less- and more-developed municipalities. That relation is analyzed in this section as Model 2 and Model 3. Model 2 concerns the more-developed municipalities of Suzhou, Wuxi, and Nanjing, where the institutional barriers are higher. The sample size for this model is 1,309, and its Nagelkerke’s R square is 0.361 (see Table 4). The sample size for Model 3 (concerning Changzhou, Yangzhou, Zhenjiang, Taizhou, and Nantong) is 1,497, with a Nagelkerke’s R square of 0.288 (see Table 4). The institutional variables associated with a Jiangsu hukou turn out to have different effects in these two models. But the demographic indicators, with the exception of age, are more or less equally strong predictors of home ownership among rural migrants. Only in Model 2 does age have a significantly positive correlation. That is probably because rural migrants need time to accumulate more capital in Suzhou, Wuxi, and Nanjing, where house prices are much higher than in the other municipalities (cf. Table 2).

Regarding the institutional factors, a Jiangsu hukou has a very positive effect on home ownership in those urban areas that have a high threshold for hukou access, namely Suzhou, Wuxi, and Nanjing. In contrast, a Jiangsu hukou is statistically not significant in the less-developed municipalities (see Table 4). Two possible explanations may be offered. The first concerns the cost of home ownership. In addition to the higher average price per sq meter for market sector housing in the more-developed municipalities, rural migrants would have to purchase a larger apartment to gain hukou access there (cf. Table 2). Given this affordability issue, access to one of the housing provident funds and to mortgage loans is a crucial precondition for home ownership. Compared to rural migrants with a non-Jiangsu hukou, workers with a Jiangsu hukou—and thus with institutional advantages and more human capital—have a greater chance to meet these preconditions. In the less-developed municipalities, where housing prices are lower, better-off rural migrants can more easily afford the house payments through capital accumulation (see Table 4).

The other explanation lies in the participation in insurance schemes. Since 2007, rural migrants with a Jiangsu hukou can transfer their pension insurance for free to other places within the same province (Jiangsu Provincial Government 2007b). It is fairly easy for an intra-province migrant to transfer participation in the social welfare scheme of his/her hometown to an urban destination within Jiangsu province. However, the situation of inter-province migrants is not that simple, and they face institutional hurdles. In that case, a better strategy for an inter-province migrant could be to first settle down in a less-developed municipality and participate in its institutions, taking advantage of its lower threshold for access. Then, after obtaining a Jiangsu hukou and a municipal hukou, it will be easier for the migrants or their children to gain access to institutions in the more-developed municipalities of the province due to the clause of free transfer. In a sense, participation in the institutions is a form of stepwise social mobility for rural migrants from a poor province. Currently, they remain at the first stage of such a strategy. The institutional barriers make it unlikely for an inter-province migrant to pursue home ownership in more-developed municipalities. In contrast, in the less-developed municipalities, where it is relatively easy for rural migrants to settle permanently, there is no obvious difference in home ownership between migrants with a non-Jiangsu hukou and those with a Jiangsu hukou.

The analysis of each municipality separately reveals more details about the inter-municipality difference in the effects of a province-level hukou and having urban pension insurance (see Table 5). For the analysis, these two were the only institutional variables we examined, due to the small size of the sample for each municipality. In Changzhou, Zhenjiang, Yangzhou, Nantong, and Taizhou, a province-level hukou did not prove to be statistically significant (or its correlation with home ownership was very weak, as in Nantong), whereas the effect of urban pension insurance was relatively strong. In contrast, both a Jiangsu hukou and pension insurance showed a positive correlation with rural migrants’ home ownership in Nanjing, Suzhou, and Wuxi. In that light, our second hypothesis appears to hold: in the less-developed municipalities, where the threshold for access to institutions is lower, the effects of institutional factors are less significant than in the more-developed municipalities, where the hukou comes with more valuable entitlements and access to it remains consequently more restrictive.

Table 5 Binary logistic regression models of rural migrants’ home ownership for each city
Full size table

5 Conclusion

This paper has explored the constraints on attainment of home ownership by rural migrants in Jiangsu province. We have raised two research questions: What are the effects of institutions and demographic characteristics on rural migrants’ home ownership rates after the hukou reform in China? and Which differences occur between destination municipalities in the same province in this respect?

We have found that after the hukou reform, the older, more highly educated, higher-income migrants with a large family are more likely than others to own private housing. Institutional factors, notably having a Jiangsu hukou and urban pension insurance, have been shown to be positive predictors of rural migrants’ home ownership, although an inter-municipality difference has been observed. In the more-developed municipalities of Suzhou, Wuxi, and Nanjing, a Jiangsu hukou has been a prominent factor in predicting rural migrants’ path to home ownership. A Jiangsu hukou, in contrast, has turned out to be irrelevant in the less-developed municipalities of Changzhou, Yangzhou, Zhenjiang, Nantong, and Taizhou.

With respect to the theoretical implications, the analysis presented in this paper has led to deeper understanding of the hukou reform. In essence, the national government intends to promote free migration within the country to accelerate urbanization. Meanwhile, provincial and municipal governments encourage rural migrants’ contribution to local economic growth. Consequently, the institutional constraints on rural migrants’ access to home ownership at their destination are gradually being relaxed. It is interesting to note the effect of this: studies referring to the situation before the reforms emphasized the institutional constraints as a dominant factor in the explanation of homeownership rates; now, demographic and socioeconomic characteristics contribute much more. This means that the explanation models for tenure differentiation in China are staring to become more consistent with findings from research in market economies.

Furthermore, this paper has documented a divergence of these effects between municipalities with a different extent of hukou reform. Rural migrants’ access to home ownership is evidently still fettered by the institutional constraints in more-developed municipalities with higher standards for hukou access.

Of course, given the limitations of our empirical material, this quantitative analysis has not broached the topic of how rural migrants respond to the hukou reform, nor how they make a trade-off between investing in housing in their hometown and purchasing a market sector dwelling at their destination. Questions such as these will be addressed in our future publications that will be based on our extensive 2012 survey in Yangzhou.

With respect to policy implications, this paper has evaluated the practical utility of the hukou reform and explored some directions for further reform. Thus far, it has relaxed the constraints on rural migrants who seek access to the local hukou system. In particular, it has released the social welfare system from hukou constraints. These institutional changes have strengthened the rural migrants’ resolve to settle down in one of the destination municipalities permanently. At the same time, they do not have to give up their home hukou. A better-off rural migrant is more inclined to invest in market accommodation, choosing ownership tenure instead of temporary housing. This propensity is especially noticeable in the less-developed destination municipalities, where housing prices are lower and participation in institutions is relatively easy. It seems that the hukou reform has contributed considerably to rural migrants’ propensity to settle down permanently in the less-developed destination municipalities. However, it is still difficult for them to settle in the large municipalities, where the social welfare systems are more extensive. These municipalities therefore need to take further steps to relax their rules for hukou status if they wish to bind the rural migrants and thereby stimulate their economy.

Despite these institutional improvements, the hukou reform has also aggravated the disparities between rural migrants, due to China’s ingrained favoritism for the local population at the expense of migrants from far, who remain “second rate” citizens. The hukou reform implemented by local governments first benefits rural migrants from the same municipality; the next beneficiaries are migrants from the same province; last to benefit are those from other provinces. Thus, the mismatch has been effectively reduced for intra-province rural migrants, but inter-province rural migrants still experience serious mismatch problems. In order to improve their situation, provincial governments could cooperate more closely to implement cross-regional policies on housing mortgages and insurance schemes, thereby limiting the scope for discrimination against inter-province rural migrants.

The emphasis on the housing qualification suggests that the recent hukou reform has been driven by the municipal governments’ intention to boost local real estate development and economic growth. It obviously benefits the better-off rural migrants who are capable of purchasing market sector housing after years of capital accumulation. They can choose to apply for a destination hukou or instead maintain their hukou at their place of origin for the sake of retaining access to their farmland. In our view, the ideal strategy for a rural migrant would be to purchase urban housing in the destination municipality, take part in some local public amenities like urban insurance, but still maintain a hukou in their home village. That would give rural migrants the flexibility to adjust to changes in their life cycle. For instance, if a migrant needs the hukou of the destination for the education of children, he/she can apply for a local hukou by virtue of home ownership.

In the real world, however, for most poor rural migrants, the hukou system is still out of reach. In order to improve their situation, more eligibility criteria could be subjected to reform; for instance, changing the rules for the insurance qualification could give them the opportunity to attain a local hukou after participating in an urban pension scheme for a certain number of years.

In the vision the national government has set forth for China’s urbanization, more migrating farmers without a local hukou would need to settle permanently at their destination. Once there, these newcomers would surely seek to obtain the same access to public amenities as their urban counterparts. In that case, the trajectory of admitting all rural migrants to the social welfare system, independent of their hukou status, would impinge on more than the insurance schemes. The way forward is through reform, and not only in the realm of insurance. Change is also needed in public housing, education, and social assistance (guaranteeing a minimum livelihood) to provide for the rural migrants or their children who will be living in the destination municipalities in the future.