Quality-Adjusted Agricultural Land Abundance Curse in Economic Development: Evidence from Postreform Chinese Panel Data

ABSTRACT

China’s family responsibility system, introduced in 1979, resulted in equal distribution in land, allowing identification of how land abundance affects development. We measure land abundance as quality-adjusted farmland per capita. We find robust evidence that higher quality-adjusted farmland per capita has a significant negative effect on growth, even after controlling for land quality and population density. Therefore, quality-adjusted agricultural land abundance confers a type of “resource curse,” which elucidates an important causal determinant of the contemporary substantial differences in the standard of living across Chinese provinces.

KEY WORDS:

• quality-adjusted farmland abundance curse
• quality-adjusted farmland per capita

Acknowledgments

The author is grateful to three anonymous referees for comments that substantially improved this article. The author also thanks Paul Beaudry, Ashok Kotwal, and seminar participants at the Canadian Economics Association Annual Meeting in Vancouver, the UBC macrointernational seminars, and the UBC TARGET student workshop for comments. The author also thanks the China Academy of Sciences’ Natural Resources Database for the agricultural and weather data.

Notes

1. In this article, industrialization refers to the structural transformation in which the sectoral composition of economic activity shifts toward industry and services (see Kuznets 1966).

2. Previous empirical works on agricultural resources use agricultural value-added and find that agricultural resources have no significant effect on growth (e.g., Papyrakis and Gerlagh 2007).

3. Mehic et al. (2013) show that FDI has significant effects on growth in Southeast Europe. Kristjanpoller and Olson (2014) examine the role of export versus import in promoting growth in Latin American countries.

4. The China Agricultural Economic Statistical Yearbook (compiled by the National Bureau of Statistics of China and published by China Statistics Press, Beijing) does not provide data on province-level yearly average yield per hectare for food crops after 1996. It still provides province-level data on average yield per hectare for cereal crops. We calculate the province-level yield per hectare for food crops in 1996 as the product of province-level yield per hectare for grain crops in 1996 and the national average yield per hectare of food crops in 1996 divided by the national yield per hectare of grain crops in.

5. For example, some authors argue that geography (including climate/weather conditions) directly affects development (Bloom and Sachs 1998; Sachs et al. 1999), while others highlight many channels (i.e., past events/institutions [Acemoglu et al. 2001; Sokoloff and Engerman 2000]; agriculture and disease [Bloom and Sachs 1998]) via which geography affects development.

6. The results with either Tempvar1 or Tempdiff as an independent regressor are almost identical to those with Tempvar2 as the independent regressor.