Abstract
Little is actually known about the factors that affected the demand and supply of agricultural credit during the Republican era, largely because data collection was rare and modern econometric techniques and the computing power necessary were simply not available. The most complete assessment was reported in John Lossing Buck’s Land Utilization in China. The important questions for economists and specialists in agricultural finance revolve around how much was borrowed, for what reasons, and at what interest rates. Answers to these are partially furnished by Buck, but not in a form that is economically meaningful. In this chapter we put Buck’s recovered data to the econometric test. The objective is to recover as best as possible the economic relationship between credit demand and interest rates, by specifying two endogenous equations for credit supply and credit demand. We consider interest rates, agricultural productivity, and special farm expenditure such as weddings and funerals as drivers of demand, and loan amount, productivity, crop yield risk, and source of credit as drivers of supply. Productivity is linked to credit amount and interest rates as well as farm size, productive and market animals, hired labor, and so on as productivity factors.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Buck , J.L. (1937). “Land Utilization in China: A Study of 16,786 Farms in 168 Localities, and 38,256 Farm Families in Twenty-Two Provinces in china, 1929-1933” University of Nanjing, Nanjing China.
- 2.
Kuribayashi, S . Edt. (2007). “Restoration of Farm Survey of Rural China in 1930s and Comparison with the Present Sampling Survey of Chinese Farms”, Final Report, Tokyo International University.
Hu, Hao , Weiwei Zheng , Minjie Yu, and Funing Zhong. (2016). “Buck ’s Original Data Mining and Integrating” Mimeo, College of Economics and Management, Nanjing Agricultural University ,
Nanjing, Jiangsu .
- 3.
The reliability of the data has been explored, with positive affirmation, by Zhong, Funing , Hao Hu and Qun Su. (2007). “Reliability of John Lossing Buck ’s Survey Data-A Preliminary Test of Grain Yields” in Kuribayashi, S. Edt. (2007). Op Cit. Chapter 3.
- 4.
See Chap. 4 for a discussion of these traditional approaches to borrowing amongst farmers. ROSCA = rotating savings and credits association.
- 5.
Buck (1937). Op Cit., page 292.
Buck , J.L. (1930). “Chinese Farm Economy: A Study of 2866 Farms in Seventeen Localities and Seven Provinces in China” University of Nanjing and University of Chicago Press, Chicago Illinois, page 235.
- 6.
The data provided in Table 6.16 are our own compilation from multiple sources, but largely from the reporting in China Weekly Review, a weekly English language news magazine. All news items between mid-1928 and late 1933 were scanned for event information, including drought , famine , floods, bandits , communist activities, anti-bandit/anti-communist activities, warlord actions, anti-Japanese activities, and so on. Locations and dates were collated as best as possible to the provinces and time schedule of the survey periods as reported by Buck .
- 7.
In the 3SLS credit demand function, if there is more loan supply in the agricultural credit market, there will be more farmers conducting borrowing behavior (coefficient of 3SLS interest rate = 0.2058877, P-value = 0). However, from the nearly perfect inelastic relationship between productivity and loan demand in the second graph, productivity does not have much of a role in impacting farmers’ decision to borrow (P-value of 3SLS production = 0.103).
Important time variables are year1930h2 (Coefficient = 0.1273513, P-value = 0). Important region variables are Shanxi (Coefficient = 0.2838039, P-value = 0.049), Guangxi (Coefficient = 0.4342342, P-value = 0.055), Yunnan (Coefficient = 0.324354, P-value = 0.098), Gansu (Coefficient = −0.772441, P-value = 0), Qinghai (Coefficient = 0.4191007, P-value = 0.044) and Ningxia (Coefficient = 0.3736906, P-value = 0.013). Gansu Province does not appear in a OLS credit demand function. In the 3SLS credit supply function, loan supply is very inelastic to loan demand (Coefficient of 3SLS loan amount = 0.9349819, P-value = 0.763), which implies Chinese farmers are price takers in the agricultural credit market. And productivity also does not impact loan supply demand (Coefficient of 3SLS loan amount = −0.0043298, P-value = 0.87). Gansu (Coefficient = 4.430215, P-value = 0) is the only valid independent variable while it does not appear in the OLS credit supply function. In the 3SLS productivity function, whether the farmer borrow money has a greatly positive impact on productivity because the coefficient of Dummy Loan is 9.688896, P-value = 0.007, which means when farmers take on more loans, the higher the productivity will be. It also might imply the loan amount was mainly used for productive purposes. The coefficient of loan supply (coefficient of interest rate = −2.31383, P-value = 0.003): the more loan supply, the less productivity will be. It implies informal borrowing sources (friends, relative, etc.) takes an important role in the loan supply side, otherwise if formal credit dominates the agricultural credit market, the increase in credit supply will not decrease productivity; for the money lender, those who spared some of their income or savings to lend to other farmers, the money they lent might have been planned for productive purposes previously. Among the other basic independent variables, the statistically significant ones are Farm Area Square (Coefficient = −0.0002252, P-value = 0), Farm Area (Coefficient = 0.0922535, P-value = 0), If has hired labor (Coefficient = 3.3936, P-value = 0). Important time variables are year1930h2 (Coefficient = −3.956419, P-value = 0.018), year1931h1 (Coefficient = 18.98447, P-value = 0), year1932h1 (Coefficient = 14.28114, P-value = 0). Important region variables are Guangxi (Coefficient = −26.44501, P-value = 0), Guizhou (Coefficient = 30.75549, P-value = 0), Yunnan (Coefficient = −12.32866, P-value = 0.002), Gansu (Coefficient = 27.37746, P-value = 0), and Qinghai (Coefficient = −12.76711, P-value = 0). Guangxi appears in 3SLS while not in the OLS production equation. On the other hand, Year1930h1 (OLS: Coefficient = −27.42663, P-value = 0) appears in the OLS not 3SLS productivity equation.
- 8.
The observations of total loan for indebted farmers are targeted to those farmers who were indebted. The sample size is small (observations = 361) compared with the original dataset.
In the 3SLS credit demand function, interest rate (Coefficient = 14.53257, P-value = 0.038) and production (Coefficient = 2.888649, P-value = 0.068both positively related with loan demand. The increase in money supply will also increase money amount demanded, which implies generally speaking farmers are willing to borrow and borrow behavior is a common to be seen for Chinese farmers at that time; increase in production also increase loan demand. In my opinion, the higher production, the larger farm size will be, which will stimulate farmers’ need for money used for productive purpose. We also can see among the indebted farmers, special events like weddings (Coefficient = 0.5694614, P-value = 0) and funerals (Coefficient = 0.2014051, P-value = 0) will increase their credit demand while birth of sons reversely decreases the credit demand . All time variables are statistically insignificant. All region variables are Hebei (Coefficient = −75.53818, P-value = 0.095), Liaoning (Coefficient = −124.3306, P-value = 0.06), Yunnan (Coefficient = 212.2831, P-value = 0.019) and Gansu (Coefficient = −249.4231, P-value = 0.02). However, Yunnan does not appear in the OLS credit demand function. Secondly, all basic independent variables and time variables are statistically insignificant in 3SLS credit supply function. Loan demand does not impact loan supply, which verifies the inference the farmers are price takers and the higher production, the lower loan supply will be. Important region variables are Hebei (Coefficient = 5.482501, P-value = 0.004), Shanxi (Coefficient = 5.841284, P-value = 0.042), Guizhou (Coefficient = 6.89809, P-value = 0.028), Gansu (Coefficient = 15.11412, P-value = 0) and Qinghai (Coefficient = 5.524437, P-value = 0.051). However, year1939h1 (OLS: coefficient = 3.906955, P-value = 0.026) appears in OLS credit demand function instead of 3SLS function. In 3SLS productivity function, Loan demand does not have relationship with production. However, the more loan supply, the lower production will be. The other basic independent variables are all statistically insignificant. The only valid time period is year1932h1 (coefficient = 22.24962, P-value = 0.094). Important region variables are Hebei (Coefficient = 17.02094, P-value = 0.076) and Gansu (Coefficient = 54.7489, P-value 0.022).
- 9.
The sample size is still 315. Targeted survey takers are farmers who borrowed for consumption purposes. The results of the 3SLS credit demand function are similar to Table 6.20. Total Loan for indebted farmers with Time and Region Variables, interest rate (Coefficient = 19.12231, P-value = 0.002) and productivity (Coefficient = 3.771076, P-value = 0.026) both positively related to loan demand. That is to say, similar to Total Loan Version, loan demand increases with loan supply; loan demand also increases with productivity. We can also see among the indebted farmers, special events like weddings (Coefficient = 0.4581681, P-value = 0) and funerals (Coefficient = 0.1939465, P-value = 0) will increase their credit demand while the birth of a son reversely decreases the credit demand . Important time periods are year1930h1 (Coefficient = −258.2811, P-value = 0.018), year1931h1 (Coefficient = −276.7796, P-value = 0.005) and year1932h1 (Coefficient = −336.8034, P-value = 0.001). Important region variables are Hebei (Coefficient = 179.4723, P-value = 0.031), Shanxi (Coefficient = 170.5514, P-value = 0,007), Guangxi (Coefficient = 331.1812, P-value 0.001), Guizhou (Coefficient = 207.7009, P-value = 0.001), Yunnan (Coefficient = 551.327, P-value = 0), Qinghai (Coefficient = 195.6869, P-value 0.006), and Ningxia (Coefficient = 211.5881, P-value 0.002). Guizhou province appears in the 3SLS credit demand function not in the OLS function while time variable year1929h1 appears in the OLS not the 3SLS . Similarly, in the credit supply function, loan demand and production cannot affect loan supply. All basic independent variables and time variables are statistically insignificant. The only two ones are region variables Hebei (Coefficient = 5.177187, P-value) and Gansu (Coefficient = −4.059061, P-value = 0.018). In the 3SLS productivity function, productivity has no relationship with loan demand while it has a negative relationship with loan supply, which verifies that Chinese farmers are price takers.
The other basic independent variables are all statistically insignificant. Important time variables are year1930h1 (Coefficient = 52.89896, P-value = 0.04), year1931h1 (Coefficient = 72.37668, P-value = 0), and year1932h1 (Coefficient = 79.37669, P-value = 0). Important region variables are Hebei (Coefficient = −37.5833, P-value = 0.053), Shanxi (Coefficient = −44.35667, P-value = 0.004), Guangxi (Coefficient = −77.4175, P-value = 0), Guizhou (Coefficient = −40.05429, P-value = 0.004), Yunnan (Coefficient = −76.22105, P-value = 0), Qinghai (Coefficient = −49.96959, P-value = 0.001), and Ningxia (Coefficient = −50.81258, P-value = 0). Among these Yunnan only appears in the 3SLS productivity function not the OLS equation.
- 10.
The sample size is only 58. Targeted survey takers are farmers who borrowed for productive purposes. In this case, the higher the supply of money (Coefficient of interest rate = −55.33578, P-value = 0.053), the less demand for productive purposes. On the other hand, the higher productivity is (Coefficient = 5.440612, P-value = 0.003), the greater the demand for productive loan. Farm size and efficiency might play a role in this part. Wedding (Coefficient = 0.130652, P-value = 0.011) and Birth of Son (Coefficient = 1.049963, P-value = 0.057) are significant special expenditures . There is no valid time and region variable in the credit demand function. In the credit supply function, loan demand (coefficient = −0.0034321, P-value = 0.053) has a slightly negative impact on loan supply and the higher productivity (coefficient = 0.0387461, P-value = 0.064), the more loan supply will be. It makes sense because productivity means higher income and higher ability to provide loans. Important time variables are year1931h1 (Coefficient = −2.336454, P-value = 0.003) and year1932h1 (Coefficient = −2.511676, P-value = 0.05). Important region variables are Hebei (Coefficient = 2.623176, P-value = 0.007), Shanxi (Coefficient = 3.984983, P-value = 0.002), Guangxi (Coefficient = 4.598918, P-value = 0.003), Yunnan (Coefficient = 4.121536, P-value = 0.009), Gansu (Coefficient = 5.078707, P-value = 0.001), Qinghai (Coefficient = 5.285053, P-value = 0), and Ningxia (Coefficient = 2.494584, P-value = 0.078). In the 3SLS production function, productivity does not have a relationship with loan demand as well as supply which implies income and savings may first be used for productive purposes which ensures productivity. The sample size is too small to make the results convincing.
Among basic independent variables, the only valid one is If has hired labor (Coefficient = 4.310013, P-value = 0.096). The only important time period is year1930h2 (Coefficient = −9.972679, P-value = 0.059). Important region variables are Guangxi (Coefficient = −39.36868, P-value = 0.079) and Guizhou (Coefficient = 41.34098, P-value = 0.041).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2018 The Author(s)
About this chapter
Cite this chapter
Fu, H., Turvey, C.G. (2018). Estimating the Demand for Farm Credit in the Republican Era. In: The Evolution of Agricultural Credit during China’s Republican Era, 1912–1949. Palgrave Macmillan, Cham. https://doi.org/10.1007/978-3-319-76801-4_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-76801-4_6
Published:
Publisher Name: Palgrave Macmillan, Cham
Print ISBN: 978-3-319-76800-7
Online ISBN: 978-3-319-76801-4
eBook Packages: Economics and FinanceEconomics and Finance (R0)