There is an interesting working paper out from the Hong Kong University of Science & Technology, “China’s mobility barriers and employment allocations“, that attempts to redirect the debate over China’s household registration (hukou) system.

China’s hukou system is intended to restrict and redirect urbanization, and this is exactly what it does: because people born with a rural hukou cannot claim government benefits in an urban area, fewer people migrate from rural to urban areas than would otherwise be the case. Most of the discussion over the hukou system (such as in these two excellent recent WSJ articles) emphasizes how it restricts migration through its links to various social services and social welfare policies. The authors of this paper (Rachel Ngai, Christopher Pissarides, Jin Wang) compare this aspect of hukou policy with a less-studied one: the potential for rural people to lose their land rights if they migrate to urban areas.

In our view, barriers to mobility from the hukou registration system arise mainly along two dimensions. First and foremost in the use of land, which is provided free by the state to rural families but is in principle withdrawn when the farmer gives up agricultural employment to move to a different job; and second in the provision of social services such as education and health, which are conditional on each person’s hukou registration and in particular the area that she lives. …

We show that the land policy embedded in the hukou system slows down migration from the land and calculate that this has led to overemployment in agriculture of 6.3 percentage points. The 6.3 points of overemployment in agriculture have come at the cost of 4.1 points underemployment in urban sectors and 2.2 points underemployment in rural non-agricultural sectors. The policy followed with respect to social transfers has further held back migration out of agriculture by another 0.4 percentage points. The biggest impact of the social transfer policy, however, is on urbanization (rather than industrialization). Because of it, rural businesses overexpanded at the expense of urban businesses, as agricultural workers prefer to stay in rural areas to benefit from their local hukou registration than move to the city and lose their local hukou. …

We find that land policy and the absence of property rights for farmers are the main channels through which the hukou system distorts both urbanization and industrialization. The social subsidies are too small by comparison, and although they have an impact on urbanization and the growth of rural enterprises, their impact on industrialization is much less. This is an important finding in light of the literature that highlights the role of hukou in restricting the access of migrants to the social services received by urban hukou holders. Such restrictions may have important social consequences but their distortionary effects on migration flows are not large.

This quantitative finding suggests that the economic benefits to China of liberalizing rural land rights would be much greater than the expansion of social welfare benefits. The theory is that if farmers had full property rights in their land, they would be able to transfer their land to others and could enjoy an income stream from the land even while working in the city. In that case more farmers would seek more productive non-agricultural employment, and national incomes would be higher as a result.

The quantitative estimate in the paper however assumes that no transfers of rural land are taking place in China today:

In the formal modelling that follows we deal with the complex issue of transfers by focusing on two extremes, one with no transfers and one with full property rights with transfers. We consider the former to be a much closer description of the present situation in China, whereas the latter corresponds to a policy reform that involves the privatization of land, something not yet contemplated by the People’s Republic. …

Although it is apparent that some unrecorded transfers of land are taking place, we ignore them in this derivation. They are small in number, the rental is unrecorded but believed to be below market rate and they do not make much difference to farmer’s attitudes. But given that some migrant workers do find ways to retain some income from their land, the impact that we calculate should be treated as an upper bound for the costs of the policy.

I’m not so sure that assumption is so easy to make. Transfers of land among farmers are allowed under current law, and have been officially encouraged (albeit with some limits) for several years now. Official statistics show that about one-third of the farmland managed under the household responsibility system has been transferred in one way or another, which is a lot more than zero.

I do think there are large economic effects from China’s rural property rights system, and the paper is right to focus on them. But the size of those effects is likely to be smaller than the estimates in the paper, as there has already been some gradual change in the direction it recommends.