The household finance literature typically ignores household migration decisions and how such decisions interact with financial conditions. We find that a relaxation of borrowing constraints can facilitate household migration to higher-tier cities where borrowing constraints are more binding than in cities of origin. Such endogenous location upgrading amplifies the increase in household housing expenditures following the relaxation of borrowing constraints, as well as intercity home price disparities.

Household credit conditions and housing boom-bust cycles

More than a decade after the Global Financial Crisis of 2007–2009, a consensus has emerged around the crucial role of household credit conditions in the housing boom-bust cycle (Mian and Sufi 2009). Yet there is less consensus on the magnitude of this effect: Favilukis et al. (2017) find that a relaxation of financial constraints substantially raised home prices, while Kaplan et al. (2020) find only a minimal effect.

These quantitative studies overlook an important household decision: migration. Credit conditions affect migration because when purchasing a home in the destination city, households face a limit on how much they can borrow to finance the purchase. This borrowing limit is more binding for those with less savings, as they will have to dramatically cut consumption and thus gain less from living in the destination city. In equilibrium, less wealthy households will sort into cities with lower home prices.

Incorporating migration into the analysis matters because it amplifies how policies or market innovations that ease household borrowing constraints can boost housing expenditures. For instance, higher borrowing limits or cash transfers will encourage households to migrate to higher-price cities. Since borrowing constraints bind more in these destination cities compared to their cheaper cities of origin, households’ marginal propensity to spend on housing is higher. Moreover, higher-price cities typically offer higher wages, which further increases housing spending.

The shantytown renovation program: A large-scale natural experiment

Between 2016 and 2020, China’s housing market experienced an average price growth of more than 40%—a surge comparable to the entire US housing boom of 2002–2006 (Mian and Sufi 2011). Unlike the US housing boom in which household credit conditions changed largely due to expanded borrowing limits, for China’s housing boom of 2016–2020, a large-scale government initiative—the cash-based shantytown renovation program—did much of the same work by converting the illiquid shanty houses into cash.

Shantytown renovation—a common urban development policy involving resettlement and infrastructure upgrades—has been central to China’s urbanization agenda since 2013. Previously, displaced residents received replacement housing as compensation. But in 2015, facing ballooning housing inventories in lower-tier cities, Beijing shifted strategy: residents instead received cash payments, with which they could purchase houses on the open market. Between 2015 and 2018, the China Development Bank (CDB) extended roughly 4 trillion RMB in loans to local governments to fund these cash payments, financed by the People’s Bank of China’s Pledged Supplementary Lending facility.

Due to the lack of formal title deeds and proper public facilities, shanty houses were deeply illiquid — they could barely be used as collateral and could not be sold in secondary markets (He et al. 2019). The cash resettlement program effectively converted these frozen assets into portable wealth—and, crucially, that wealth could be deployed anywhere.

Empirical finding on home prices and migration

Using China’s shantytown renovation program as a natural experiment, we find strong evidence that the cash transfer, by relaxing household borrowing constraints, can induce household migration to higher-tier cities and raise home prices there.

We first combine proprietary CDB loan data with National Population Census data and construct two measures: a measure of cash resettlement in loan originating cities, and a Bartik-style measure of cash inflows received in destination cities via preexisting migration networks.

The results show that cities with more cash distributed to their residents saw no acceleration in housing price growth after 2015. By contrast, cities that received cash inflows through the migration network experienced robust and sustained housing price appreciation, with new housing supply also rising modestly. These results suggest that the program fueled housing demand relative to supply not where the money originated, but where migrants chose to spend it.

Critically, this destination-side effect is far stronger along city pairs with a larger home price gap—that is, the price-appreciation effect of inflows originating from lower-priced cities is roughly three times that of cities with prices similar to the destination city. This pattern is precisely what one would expect if cash resettlement relaxed borrowing constraints that were preventing households from purchasing homes in more expensive cities.

We then use the population census data to identify two migration channels underlying the cross-city effect of cash resettlement. The first channel is “downgrade mitigation”—among households that had already migrated to expensive cities but kept shanty property in their city of origin in 2015, cash transfers in their city of origin significantly raised their likelihood of staying and becoming homeowners in the expensive destination city by 2020. The effect is statistically significant and economically large for urban migrants but absent for rural migrants (who were not targeted by the program), ruling out generic city-pair economic factors. The second channel is “upgrade facilitation”—among cash-receiving households, we find that roughly 22% migrated to other cities by 2020; these individuals were predominantly of working age and cited employment as the primary motive for migration. Crucially, both channels are significantly stronger between city pairs with large house-price gaps, consistent with the larger price-appreciation effect of inflows from lower-priced cities.

Quantitative finding: Migration amplifies household housing expenditure by 40%

To quantify the amplifying role of migration, we build a dynamic spatial general equilibrium model with endogenous migration and collateral constraints. The model features an important distinction from the existing literature (Bilal and Rossi-Hansberg 2021, Giannone et al. 2023, Greaney 2023): households’ residence location and homeownership location can differ, as households often rent in their destination city while retaining property in their city of origin. We calibrate the model to five aggregate locations and solve for the household forward-looking consumption-saving, home purchase and location choices from 2015 onward.

To isolate the effect of migration, we utilize the “voucher resettlement” scheme (“房票” in Chinese) to investigate a counterfactual policy, where the voucher can be spent only on local home purchases, thereby shutting down migration-induced housing spending. Voucher resettlement was adopted in few cities such as Xi’an and Jining in 2016.

The results are striking. Under cash resettlement, the average housing expenditure multiplier — the increase in home purchase spending per yuan of transfer—is 1.14. Under voucher resettlement, it falls to 0.80. Migration thus amplifies the household housing expenditure response by 42.5%. In equilibrium, absent the program, average home prices across China would have been around 4% lower in 2016–2020; this price effect is entirely attributable to the mobility channel, as voucher resettlement generates virtually no aggregate price impact. Our model explains roughly 20% of detrended home price growth over this period.

Policy implications

Our analysis carries two important messages for policymakers. First, researchers typically associate higher leverage in the city of origin with more binding borrowing constraints. Our analysis suggests that this can be misleading. A household may appear relatively unconstrained today precisely because the constraint has already pushed them away from the expensive city where they would prefer to live. Their true borrowing constraint is more accurately described by what their leverage would be in their preferred destination city, rather than their leverage in their city of origin. If households with modest savings are more likely to upgrade locations with the transfer, then policies that aim to stimulate household spending should target these households, rather than poorer or richer households.

Second, local stimulus policies may enlarge spatial inequality in fiscal capacity. When the cash-receiving households migrate to higher-tier cities and purchase homes there, local governments in the higher-tier cities benefit from the booming housing market. However, the cost of the stimulus is still borne by the cash-distributing governments. Household migration essentially leads to a fiscal transfer from lower-tier to higher-tier cities. The voucher transfer that limits non-local spending will be preferred if policymakers are concerned about fiscal inequality across cities.

Zhiguo He, Stanford University and National Bureau of Economic Research; Zehao Liu, Renmin University of China; Xinle Pang, University at Buffalo, SUNY; Yang Su, Chinese University of Hong Kong; Kunru Zou, Hong Kong Baptist University

References

Bilal, Adrien, and Esteban Rossi-Hansberg. 2021. “Location as an Asset.” Econometrica 89 (5): 2459–2495. https://doi.org/10.3982/ECTA16699.

Favilukis, Jack, Sydney C. Ludvigson, and Stijn Van Nieuwerburgh. 2017. “The Macroeconomic Effects of Housing Wealth, Housing Finance, and Limited Risk Sharing in General Equilibrium.” Journal of Political Economy 125 (1): 140–223. https://doi.org/10.1086/689606.

Giannone, Elisa, Qi Li, Nuno Paixão, and Xinle Pang. 2023. “Unpacking Moving: A Quantitative Spatial Equilibrium Model with Wealth.” Bank of Canada Staff Working Paper No. 2023-34. https://doi.org/10.34989/swp-2023-34.

Greaney, Brian. 2023. “Homeownership and the Distributional Effects of Uneven Regional Growth.” Social Science Research Network Working Paper No. 4406024. https://dx.doi.org/10.2139/ssrn.4406024.

He, Shenjing, Dong Wang, Chris Webster, and Kwong Wing Chau. 2019. “Property Rights with Price Tags? Pricing Uncertainties in the Production, Transaction, and Consumption of China’s Small Property Right Housing.” Land Use Policy 81: 424–33. https://doi.org/10.1016/j.landusepol.2018.10.038.

He, Zhiguo, Zehao Liu, Xinle Pang, Yang Su, and Kunru Zou. 2026. “Household Migration and Collateral Constraint: Cash-based Housing Resettlement in China.” National Bureau of Economic Research Working Paper No. 34982. https://doi.org/10.3386/w34982.

Kaplan, Greg, Kurt Mitman, and Giovanni L. Violante. 2020. “The Housing Boom and Bust: Model Meets Evidence.” Journal of Political Economy 128 (9): 3285–3345. https://doi.org/10.1086/708816.

Mian, Atif, and Amir Sufi. 2009. “The Consequences of Mortgage Credit Expansion: Evidence from the U.S. Mortgage Default Crisis.” Quarterly Journal of Economics 124 (4): 1449–96. https://doi.org/10.1162/qjec.2009.124.4.1449.

Mian, Atif, and Amir Sufi. 2011. “House Prices, Home Equity-Based Borrowing, and the U.S. Household Leverage Crisis.” American Economic Review 101 (5): 2132–56. https://doi.org/10.1257/aer.101.5.2132.