Jingyi Fang

We investigate the role of reduced migration barriers in enhancing allocative efficiency using China's unified hukou reform in the 2000s. According to existing literature, firms' factor productivity is dispersed due to heterogeneous input distortions. Higher labor productivity indicates that the firm was initially more distorted in the local labor market. Our findings reveal that hukou reform disproportionately favored more productive and, consequently, more distorted firms. More specifically, while less productive firms reduce employment by 7.4% after the reform, more productive firms hired 29% more workers relative to them. The increased labor productivity of less productive firms and the decreased labor productivity of more productive firms suggests a more concentrated distribution of productivity and reduced distortions. Upon examining other pre-reform labor regulations and post-reform city-level competitions, we observed that the gains in allocative efficiency are more pronounced in areas with stricter labor regulations but are mitigated in cities facing with higher labor competition. As expected, non-state-owned enterprises, labor-intensive firms, and larger firms were more affected by the reform. When integrating the reduced-form estimated coefficients into aggregate analysis, our results show that the hukou reform increased aggregate productivity by 4.5% and expanded aggregate output in the manufacturing sector by 19% for reformed cities.

Canonical life-cycle models predict that rational, fully forward-looking agents should perfectly smooth consumption over their lifetime. This prediction, while not supported by the data, has been mostly tested for developed countries like the U.S. We use recently available, rich longitudinal data for a large sample of households to understand patterns of consumption expenditures, income growth, and savings rates in India. Our empirical analysis has several important findings. First, growth in total household consumption and income is comparable to that of the U.S. However, unlike the U.S., Indian households exhibit no growth in non-durable consumption expenditures after adjusting for family size. We document significant heterogeneity along with various population sub-groups, but none of them exhibit growth close to U.S. households. Savings rates, measured as total income net of non-durable expenditures, on the other hand, exhibit a strong hump over the life-cycle. We present evidence that the need to save for lumpy investments such as housing, cars, tractors, and cattle are key drivers of the high savings rate growth over the life-cycle for Indian households.

After the hukou policy is relaxed and the commercial housing replaced the housing assignment program in China in the beginning of 2000s, top tier cities (Beijing, Shanghai, Guangzhou, Shenzhen) experienced a large migration wave, and college workers prefer top tier cities even more. To find out the reason behind these facts, we use a simplified structural model with general method of moment for estimation, and found that there is a huge differential in preference for hukou and non-hukou workers, while non-hukou people are mostly attracted to high wage rate in the destination cities, hukou people are willing to forfeit a higher wage rent, get a stable job, and enjoy local amenities, some of which come with their local hukou registration. There is a differential pattern between college and non-college workers as well, while college workers are paid well, value their wages, and don’t care about the rentals, non-college workers are spending 75% of their income on rent but value local amenities in their residential cities a lot. While higher fraction of college educated workers elevates the amenity score, migration flood, especially low educated labors, will dilute and exhaust the local amenities.

Does consumption smoothing of the elderly across time and individuals indicate the efficiency of the overall insurance mechanisms? What is the extent of suboptimality in market and non-market insurance? To answer these queries, we employ China Health and Retirement Longitudinal Study (CHARLS) panel data covering 2011, 2013, 2015, and 2018 and examine the effectiveness of both institutional and personal insurance channels for elderly individuals in China. While the conventional tests reasonably support consumption smoothing across time and individuals against adverse health shocks, especially for basic food consumption, our results show that the welfare costs associated with such health shocks are not necessarily small, especially in rural areas. These findings call for strengthened long-term care and pension systems as well as other social safety nets, which can be welfare-enhancing even when consumption is not sensitive to shocks.

Using a novel panel survey, we show that households in India build a stock of physical as- sets over their life-cycle — an 80% growth at its peak. Saving for these lumpy asset purchases comes at the cost of crowding out non-durable consumption. Unlike the US, family adjusted non-durable consumption in India has a flat profile that is neither due to smoothing nor subsis- tence income. Using reported intentions of durable purchases, and event studies around actual purchases, we also show that future durable purchase motive indeed increases the household savings rate — anywhere between 4 and 10% in the case of housing.