Integrating China’s 200 million rural-to-urban migrants into urban society is a critical
challenge that, if unsuccessful, could undermine the entire urbanization project. To this
end, understanding and responding to migrants’ housing needs, goals, and difficulties is
an important aspect of ‘successful’ urbanization. Doing so is difficult, however, because
of the complex legacy of housing reform and the transitional state of the housing market,
and because so little is known about migrants’ housing preferences and behavior.
This paper fills some of the gaps regarding migrant housing choice, demand, and quality
using data from a purpose-designed survey of 800 low-status migrants in Tianjin. Results
show that in many cases these individuals do not to exercise housing ‘choice’ as much as
they undergo housing ‘sorting’ that follows from occupational choices. That is, less than
half of our respondents got their housing through the private rental market and only a
slightly higher share pay any rent. Employment variables (industry sector, employer type)
are consistently and strongly significant across our housing choice models and
significantly affect housing quality as well. Nonetheless, a low-cost rental sector does
exist, serving about two-fifths of migrants in our sample. Within this subset, housing
demand is more consistent with theory in the sense that income and life cycle factors are
important and the role of employment characteristics is diminished.
In all models individual migration characteristics, such as duration of residence, future
migration plans, and sending remittances home, are significant, though which particular
characteristics matter varies. We take this as an indication that migration status affects
housing outcomes in multiple and subtle ways. This perspective differs, somewhat from
the literature on housing choice in urban China, which emphasizes the role of institutional
factors in determining cross-group housing outcomes. Although our results do not
directly contradict these claims, our findings of (1) substantial variation in the
determinants of housing choice/demand within the migrant pool, combined with (2) the
‘sorting’ of migrants into housing based on employment choices, suggests that at least
some of the differential in housing outcomes between migrants and other urban groups is
a result of individual migration characteristics and employment choices rather than
institutional factors in the housing market.
Ultimately, we read the empirical results of our study as indicating that the primary
policy prescription of the urban China housing choice literature – to eliminate residual
barriers that prevent migrants from accessing low-cost public-sector rentals is insufficient
and may not respond to the concerns of migrants themselves. Of course, removing such
barriers would not be an unwelcome step but, for instance, it is unlikely to have any
impact on the housing situations of the half of all migrants that obtain housing through
their employer. In short, a range of policies will be necessary to support the housing goals
of migrants who have different housing needs and face different constraints in meeting
them.