The impact of non-farm activities on agricultural investment in Vietnam: the difference made by credit constraints levels

Credit is a necessary and important factor in the agricultural production process in many poor countries. Credit access can considerably increase the financial ability of farm households for agricultural inputs and productive investments in both the short-run and long run (Lin et al., 2019). Indeed, credit has been considered as one efficient way to improve agricultural productivity and reduce poverty. In developing countries, particularly in rural areas where the credit markets are imperfect, farmers cannot easily access credit sources. With constrained credit, rural households have difficulties in making agricultural inputs investment and consequently must limit their production and smoothing consumption (Oseni and Winters, 2009). Some empirical literature has found that in rural areas of developing countries, credit constraints have significant adverse effects on agricultural productivity (Guirkinger and Boucher, 2008; Dong et al., 2012) and farm investment (Carter and Olinto, 2003). One way farmers used to overcome the constraints and imperfections of the credit market is by diversifying their livelihoods into non-farm activities (Oseni and Winters, 2009). The literature indicated that the income source obtained from non-farm activities could help farm households for improving their household income (Ferreira and Lanjouw, 2001; Nnadi et al., 2020; Kumar et al., 2020), smoothing consumption (Seng, 2015; Mishra et al., 2015; Abdurezak and Ahmed, 2020), and reducing poverty (Haggblade et al., 2010; Hoang et al., 2014; Bui and Hoang, 2020). Sometimes, farm income is not sufficient to supply a sufficient livelihood (Minot et al., 2006), which can be a push factor driving rural households to seek opportunities for employment outside farm activities. In farming activity, farm households always face many risks or limited riskbearing capacity, inducing household members to engage in non-farm activities to reduce risk and reduce consumption uncertainties (Reardon, 1997; Barrett et al., 2001; Oseni and Winters, 2009). In addition, the non-farm income can be used in agricultural production when farmers do not have enough financial capacity to pay for farm inputs. Thus, participation in the non-farm market could help farm household to relax the liquidity constraints they face when credit is not available (Pfeiffer et al., 2009). Both the loss of family labour in the shift from farm activity to non-farm activities and access to non-farm income can influence agricultural production in direct and indirect ways. Besides the labour-lost effect, the earnings from non-farm employment can provide cash for farmers to make investments in agricultural production to enhance productivity. These investments could be for the short term such as the purchase of fertilisers, feed, herbicide, pesticide, and other inputs, or for the long term such as investments in machinery or irrigation or the adoption of new technologies (high yielding seed or improved seed) (Pfeiffer et al., 2009). There are numerous studies showing the relationship between non-farm activities and agricultural inputs investment/expenses in several developing countries. The findings reveal this effect can be positive, negative, or nil (equal to zero), depending on the context. The positive relationship between non-farm participation and the purchasing of productive agricultural assets has been explored in many countries, including Bulgaria, Nigeria, Mexico, Philippines, and Ghana (Hertz, 2009; Oseni and Winters, 2009; Pfeiffer et al., 2009; Takahashi and Otsuka, 2009; Anríquez and Daidone, 2010). These studies all concluded that the income from nonfarm activities could loosen the credit constraints for farm households who were now able to pay more for inputs in agricultural production. However, in Albania, Kenya and China, non-farm employment had been found to have a negative impact or no impact (Albania: Kilic et al., 2009; Kenya: Mathenge et al., 2015; China: Huang et al., 2009) on agricultural input expenditures. Ahituv and Kimhi (2002) analysed farm households’ off-farm labour and farm capital investment decisions and found a strong negative associaHang Thi Thuy NGUYEN*, Xuan Hung PHAM* and Takumi KONDO**