Does land certification increase farmers’ use of organic fertilizer? evidence from China

ABSTRACT

Based on the data from the China Rural Household Panel Survey, a panel logit model is adopted to examine the impact of land certification on farmers’ use of organic fertilizers. The results reveal that land certification increases the probability of farmers using organic fertilizer by approximately 15%, with an average expenditure and amount of 35 yuan per mu and 23 kg per mu, respectively. This is mainly achieved by stabilizing farmers’ expectations of property rights and increasing the availability of agricultural subsidies. Moreover, it varies significantly with different production characteristics, property rights, and regional groupings. Specifically, land certification mainly benefits small-scale farmers planting food crops, farmers with contracted land or without land adjustment experience, and farmers in western regions, but it is not beneficial for others. The study highlights the role of land certification in improving the quality of cropland and is a reference for China and other developing countries.

KEYWORDS:

• Land certification
• panel logit model
• organic fertilizer
• land conservation
• land system

Introduction

In recent years, the rapid growth of the global population has significantly increased the demand for food (FAO, UNICEF, WFP, & WHO, 2021). Food security, one of the crucial components of the sustainable development goals (SDGs) of the United Nations, has aroused broad concerns in the world (Aryal et al., 2021). Especially in China, which has the largest population but scarce cultivated lands, nothing is more critical than ensuring food security and feeding more than 1.4 billion people (Fan et al., 2012). As one of the most critical measures to increase food production, the importance of improving cultivated land quality is self-evident (Belete, 2022; van Wesenbeeck et al., 2021). However, the existing quality of China’s cultivated land does not usher optimism (Fan et al., 2012; Guo, 2012; Zheng et al., 2021). In the past few decades, due to rapid industrialization and urbanization and the overuse of chemical fertilizers and pesticides, the quality of cultivated land in China has declined sharply (Chen et al., 2022; Fan et al., 2012; Ma et al., 2018). According to the 2019 National Bulletin on the Quality of Cultivated Land released by the Chinese government, the area of excellent cultivated land in the country is only 632 million mu (1 mu = 1/15 hectare), accounting for only about 30% of the total cultivated land. This indicates that a large amount of cultivated land is still at a low-to-medium level, and high-quality cultivated land is relatively scarce.

Worldwide farming practices have proved that organic fertilizer is one of the most effective measures to improve the quality of cultivated land and realize the sustainable utilization of cultivated land (Belete, 2022; King, 1911; L. Gao et al., 2019; Zheng et al., 2021). Organic fertilizers are natural sources of minerals containing appropriate amounts of essential plant nutrients (Aryal et al., 2021; Shaji et al., 2021). It has a wide range of advantages: (i) improves soil properties and soil structure, thereby improving land quality; (ii) helps to replenish lost organic matter in the short and long term, thereby making soil fertility last for four to five years; (iii) enhances existing soil nutrients and improves soil health, resulting in healthy crop growth with minimal nutrient density; and (iv) minimizes land degradation and environmental damage, and achieves sustainable agricultural development without reducing crop yields (Chu et al., 2012; Jacoby et al., 2002; Li & Shen, 2021; Shaji et al., 2021). Therefore, ways to improve the quality of cultivated land and maintain national food security by increasing the use of organic fertilizer by farmers are increasingly important.

Several studies have discussed the determinants of farmers’ use of organic fertilizer. For example, Belete (2022) and Abebe and Debebe (2019) illustrated that in Ethiopia, gender, health status, education level of the head of household, family size, livestock units, land deed, and farm size were positively correlated with the use of organic fertilizers; whereas, soil fertility, off-farm income, credit access, and marital status of the head of household were negatively correlated with it. According to the study of Aryal et al. (2021) on the Indo-Gangetic Plains in India, Nepal, and Bangladesh, gender, education, migration, access to markets, training, off-farm income, and wealth were key factors influencing farmers’ adoption of organic fertilizers. In addition, evidence from China shows that education level, planting years, training, and land size increased the probability of farmers adopting organic fertilizers, while the distance to the market and farmers’ economic cognition decreased it (Chen et al., 2022). Nevertheless, relatively few studies explore the impact of land tenure system reform on farmers’ use of organic fertilizers (Huntington & Shenoy, 2021), despite its profound impact on rural economic development (Deininger et al., 2019).

Land certification forms the core of the reform of the land tenure system. It is a prerequisite for the optimal allocation of land resources, and an institutional guarantee for adequate protection of land property rights (De Soto, 2000). Since the 1950s and 1960s, land certification reform has been widely carried out by governmental organizations and international development agencies (de Janvry et al., 2015). Since the issuance of the Land Management Law of the People’s Republic of China in 1998 and the Land Contract Law of the People’s Republic of China in 2003, China has launched several reforms in land certification. In 2013, marked by Document No.1 of the Central Committee of the Chinese Communist Party (CPC), the Chinese government carried out a new round of land certification reform that required five years to complete. According to the latest statistics from the Ministry of Agriculture and Rural Affairs, by the end of 2020, the completion rate of rural land certification exceeded 96%, and more than 200 million Chinese farmers had received land certificates.

With the development of China’s land certification reform, the relevant literature is increasingly rich. Existing studies have highlighted the impacts of land certification on land transfer, cropland abandonment, labor migration, credit availability, and agricultural productivity (de Janvry et al., 2015; Deininger et al., 2019; Holden et al., 2011; X. Gao et al., 2021; Zheng & Qian, 2022). Meanwhile, a few Chinese scholars have started to examine the relationship between land certification and farmers’ use of organic fertilizers. For example, using data from 6,864 rural households in 9 provinces in 2000 and 2008, Huang and Ji (2012) found that farmers’ access to land contracts and land certificates increased the use of organic fertilizers. Based on the survey data of 818 farmers in Guangxi Province in 2018, M.S. Qian and Lu (2020) pointed out that land certification increased the probability of farmers adopting organic fertilizer. Similarly, using a survey sample of 694 farmers in 3 provinces in 2017, Zhou and Wang (2019) concluded that land certification increased the amount of organic fertilizer used by farmers.

The above literature lays the foundation for this research but also leaves room for improvement. First, Huang and Ji (2012) mainly studied land certification before 2013, and although M.S. Qian and Lu (2020) and Zhou and Wang (2019) both used the research data after 2013, they did not use the certification time to identify whether it was the new round of land certification. Second, they failed to simultaneously examine the impact of land certification on the probability of using organic fertilizer and the amount of organic fertilizer and did not further discuss its potential mechanism and multiple heterogeneous effects. Third, their data samples were relatively small and not nationally representative, covering only some provinces in China. Finally, the above studies were written in Chinese, which limited the international dissemination of subject knowledge to a certain extent.

This study extends prior literature in the following three aspects. First, it uses certification time to identify the new round of land certification reform in China and examine its impact, mechanism, and heterogeneous effects on farmers’ organic fertilizer use. Second, it comprehensively considers the impacts of the new round of land certification on whether farmers use organic fertilizer and the expenditure and amount of organic fertilizers used by farmers. Third, it adopts nationally representative panel data from 29 provinces across the country in 2017 and 2019 and employs various panel models to yield consistent results, which may enhance the credibility of the estimates.

Literature review and theoretical analysis

Land certification is a process of proving the area, space, duration, and legal rights of the land owned by the subject of property rights through land registration. It also refers to issuing formal qualification documents consistent with the contents of the real estate register on the subject of property rights. Theoretically, it can increase farmers’ use of organic fertilizers by improving their property rights expectations and increasing the availability of agricultural subsidies.

Land certification, the expectation of farmers’ property rights, and the use of organic fertilizer

According to the theory of planned behavior, any change in behavior is a change based on subjective will (Ajzen, 2005). Land certification can promote farmers’ use of organic fertilizer mainly because it improves their property rights expectations and enhances their perception of property rights security (L. Qian et al., 2019). As a long-term investment, the use of organic fertilizer is based on land tenure security (Besley, 1995). If the farmer expects that his land rights are insecure, he is likely to give up the investment in organic fertilizer (Abdulai et al., 2011), or even implement predatory operations on the land (Zheng & Qian, 2022). This is because insecure land property rights mean that it is difficult for property owners to have full confidence in recovering the entire income stream generated by the investment (Roth & Haase, 1998).

Land certification is the legal confirmation of land rights. The new round of land certification requires that under the premise of clearly defined physical boundaries and ownership boundaries of land, the rural land contractual management certificate with legal significance should be issued to farmers. This is conducive to solidifying the property rights of farmers’ identity, enhancing the stability, security, and exclusivity of land property rights (Kassa, 2014), as well as stabilizing farmers’ production expectations (Huang & Ji, 2012). This will undoubtedly increase the subjective effect of farmers on land property rights and increase their economic incentives to invest in the land (Saint Macary et al., 2010). Therefore, a new round of land certification centered on legal protection and empowerment would help farmers establish positive property rights expectations, and guide them to increase the possibility of using organic fertilizers.

Land certification, agricultural subsidies, and farmers’ use of organic fertilizer

Agricultural subsidy is an incentive and restraint mechanism, and farmers’ enthusiasm for agricultural production is positively related to the availability of agricultural subsidies (Strijker, 2005). In the Chinese context, in addition to the subsidies for purchasing agricultural machinery and materials, other agricultural subsidies, such as direct subsidies for grain and for improved varieties, are generally based on the land area marked on the rural land contractual management certificate (Zheng & Qian, 2022). The lack of land certificates may lead to farmers’ agricultural subsidies being misappropriated or intercepted. If farmers lack the corresponding land certificate, they are less likely to obtain agricultural subsidies, which may reduce the probability of farmers using organic fertilizer.

Compared with the incomplete issuance of land certificates (Ye et al., 2000), the current land certification provides land certificates with an accurate area, clear boundaries, and clear ownership to farmers in a complete manner. It makes land certificate a basis for issuing agricultural subsidies (Zheng et al., 2020). Therefore, the complete issuance of land certificates will theoretically help increase the availability of agricultural subsidies, reduce the inaccuracy of receiving agricultural subsidies, and prevent the embezzlement or interception of agricultural subsidies by illegal persons. Consequently, this protects farmers’ legitimate land rights and interests, improving their economic incentives for long-term investment, and enhancing the possibility of farmers’ use of organic fertilizer (Saint Macary et al., 2010).

Research design

Data

The data used in this study are from the China Rural Household Panel Survey (CRHPS), a database jointly established by Zhejiang University and the Southwestern University of Finance and Economics in 2013. The CRHPS data are collected every two years from the whole country, covering 29 provinces in China except Hong Kong, Macao, Taiwan, Xinjiang, and Tibet (Figure 1). The survey uses stratified sampling and three-stage proportional to population, ensuring that the data are representative at rural, urban, provincial and national levels (W. Qian, 2020). The dataset is comprehensive and informative, involving the basic structure of rural households, income and expenditure, agricultural production, land use, and other aspects. However, the data related to organic fertilizer have been collected since 2017; hence, the most recent two years of panel data have been used for analysis.

Figure 1. Changes in farmers’ use of organic fertilizer from 2017 to 2019.

Since this study aims to explore the impact of land certification on farmers’ use of organic fertilizers, the sample of farmer households without contracted land in rural areas (loss rate: (21815–21176)/21815 = 2.93%) was deleted and further retained the following key variables: land certification, organic fertilizer, individual, economic, labor, and land characteristics, among others. Finally, large-scale panel data of 21,176 farmer samples across 29 provinces of China in 2017 and 2019 were obtained.

Variables

1. Dependent variable: organic fertilizer use. Compared with traditional farmhouse organic fertilizers such as straw and manure, commercial organic fertilizers processed from various animal wastes and plant residues have a higher cost to meet relevant national standards and can better reflect the land quality conservation behavior of farmers. Therefore, this study mainly adopts the use of commercial organic fertilizer as the dependent variable. As observed in Figure 1, from 2017 to 2019, Chinese farmers’ adoption rate of commercial organic fertilizer exhibited an increasing trend.

2. Independent variable: land certification. Drawing on the research of Zheng and Qian (2022), a dummy variable is used to represent whether the family has received the rural land contract management certificate since 2013 to reflect the new round of land certification reform in China.

3. Control variables. Drawing on related research of Zhou and Wang (2019) and Qian et al. (2020), the control variables of this study mainly include individual, economic, labor, and land characteristics. Specifically, individual characteristics include age, gender, and education level of the householder. In addition, the study uses the family’s total income to reflect the family’s economic status. It should be noted that the total household income of many rural households is negative, so the unit of ten thousand yuan is used here rather than using the logarithm method. Moreover, the present study uses the number of the agricultural labor force and the proportion of off-farm employment to reflect farmers’ human capital. The land characteristics mainly include the per capita contracted land area, average plot area, and the market value of the contracted land, which are used to reflect the family’s land resource endowment, the degree of fragmentation of the plot, and the economic value of the contracted land, respectively. The definitions and descriptive statistics of the main variables are shown in Table 1.

   Table 1. Variable definitions and summary statistics.

   Variables	Variable definitions	Obs	Mean	SD
   Organic fertilizer	Whether to use commercial organic fertilizer (yes = 1, no = 0)	21,152	0.559	0.497
   	Whether to return manure to the field (yes = 1, no = 0)	13,073	0.668	0.471
   	Whether to return straw to the field (yes = 1, no = 0)	11,040	0.397	0.489
   	Whether to use organic fertilizer (yes = 1, no = 0)	21,176	0.792	0.406
   Land certification	Since 2013, whether farmers have received the land certificate (yes = 1, no = 0)	21,176	0.555	0.497
   Age	Age of the householder	21,176	52.690	11.251
   Gender	Male = 1, female = 0	21,176	0.829	0.377
   Education	Illiteracy = 1, elementary school = 2, middle school = 3, high school = 4, technical secondary school = 5, junior college = 6, undergraduate = 7, graduate = 8	21,176	2.735	1.125
   Rural labor	Number of agricultural labor force (person)	17,729	1.485	0.968
   Off-farm ratio	Household non-farm employment ratio	21,176	0.275	0.288
   Income	Total family income (ten thousand yuan)	21,176	5.384	9.823
   Land size	Per capita contracted land area of the household (mu)	21,176	2.628	5.637
   Plot area	Operating area of each plot (mu)	20,569	2.992	16.251
   Land value	How much do you estimate the value of your contracted land (yuan, logarithm)	20,855	7.491	2.353
   Farmers’ property rights expectations	Whether land certification can bring benefits (yes = 1, no = 0)	21,176	0.893	0.309
   Availability of agricultural subsidies	Whether to receive agricultural subsidies (yes = 1, no = 0)	20,828	0.732	0.443

   1 yuan≈0.15 dollar; 1 mu = 1/15 hectare.

Estimation strategy

Considering that the use of commercial organic fertilizer is a dummy variable, the panel composed from data in CRHPS 2017 and 2019 is a two-period unbalanced panel. Therefore, available models are discrete panel data models, namely the panel logit model and panel probit model. Correspondingly, the former requires the random error term to obey a logical distribution, and the latter requires the random error term to obey a normal distribution. However, the panel logit model can choose random effects or fixed effects, while the panel probit model can only estimate random effects. Therefore, this study chooses the panel logit model and constructs the following empirical equation:

(1) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\alpha }_0+{\alpha }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\alpha }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$(1)

Here, i represents farmer households, and t represents the survey year. Organic_fert indicates whether to use commercial organic fertilizer, Certification indicates land certification, and X indicates a series of control variables. Furthermore, α₀ is the intercept term; δ_i represents the fixed effect of farmers that mainly controls for factors such as the ability and preference of farmers; λ_t represents the year-fixed effect that mainly controls for the factors related to the use of organic fertilizer, such as national policies or macroeconomic conditions, which change over time but do not change with the farmers; and ε_it is the random disturbance term.

Impact of land certification on farmers’ organic fertilizer use

Among the linear panel models, most scholars choose fixed effects by default. However, for nonlinear panel models, such as the logit model, it is generally believed that it is better to use the random effect model than the fixed effect model especially in this case because controlling a large number of individual dummy variables in the fixed effect model may lead to coefficient estimation bias. Fixed effects models generally have two complex problems to deal with: one is the loss of a large number of samples, resulting in a significant reduction in sample size, which leads to biased estimates (Holden et al., 2009); the other is that the conditional likelihood function eliminates the individual fixed effects during estimation. If the individual fixed effects are not assumed to be zero, it will be difficult for the fixed effects model to predict the probability of events. In addition, based on the Hausman test of fixed effects and random effects, this study ultimately chose the panel logit random effects model to estimate the impact of land certification on farmers’ organic fertilizer use.

Benchmark regression

Table 2 reports the stepwise regression results of land certification on the use of organic fertilizer, and the estimated coefficients are all presented as average marginal effects. The estimated results in Columns (1) to (5) show that, regardless of whether control variables are added, the estimated coefficients of land certification are always positive and significant at the 1% level, and the regression results are robust. Compared with the farmers who do not hold land certificates, the farmers who hold them have an average of approximately 15% higher probability of using organic fertilizer. These estimates are broadly consistent with existing studies (Huang & Ji, 2012; M.S. Qian & Lu, 2020). Overall, the above findings indicate that by issuing legally valid land certificates to farmers, land certification can not only stabilize farmers’ property rights expectations but also improve the availability of agricultural subsidies, which in turn will help to encourage farmers to use organic fertilizers and thus improve the quality of cultivated land.

Table 2. Estimated results of land certification on the use of organic fertilizer.

	Panel logit model
	(1)	(2)	(3)	(4)	(5)
Land certification	0.151*** (0.036)	0.150*** (0.037)	0.155*** (0.037)	0.158*** (0.041)	0.156*** (0.043)
Age		−0.006*** (0.002)	−0.005*** (0.002)	−0.002 (0.002)	−0.002 (0.002)
Gender		−0.224*** (0.050)	−0.230*** (0.050)	−0.061 (0.058)	−0.063 (0.059)
Education		0.162*** (0.019)	0.142*** (0.019)	0.096** (0.022)	0.097*** (0.022)
Rural labor			0.015*** (0.002)	0.013*** (0.003)	0.014*** (0.003)
Off-farm ratio				0.252*** (0.023)	0.254*** (0.023)
Income				−0.347*** (0.087)	−0.407*** (0.090)
Land size					−0.011*** (0.004)
Plot area					−0.001 (0.002)
Land value					0.019** (0.009)
Obs	21,152	21,130	21,130	17,689	17,326
Log likelihood	−14415.102	−14334.081	−14311.673	−12029.744	−11766.036
Wald chi2	17.27	137.16	169.90	194.85	209.25

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; To preserve the true information of the sample as much as possible, we did not unify the sample size because the explained variables in this study are not unique, but the stepwise regression results show that this does not affect the validity of the estimation results.

Robustness test

1. Changing method: use the panel probit model and the panel linear probability model (LPM). As pointed out above, for the processing of discrete binary variables of panel data, in addition to the panel logit model, the panel probit model is also a commonly used estimation method. Moreover, regardless of the accuracy of the estimated coefficients, the panel LPM is also commonly used as a compromise estimation method. To this end, this study also uses the panel probit model and the panel LPM to re-estimate the empirical Equation (1)(1) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\alpha }_0+{\alpha }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\alpha }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$(1) . Table 3 reports the estimation results of the two models. It can be found that although the estimated coefficients of land certification are different among different models, no matter which method is used to estimate, the estimated coefficients of land certification are significant. It is positive and passed the test at a significance level of 1%, which shows that land certification can indeed increase the probability of farmers using organic fertilizers and help improve land quality.

Table 3. Land certification and use of organic fertilizer: Robustness test I.

	Panel probit model				Panel LPM model
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Land certification	0.090*** (0.022)	0.093*** (0.022)	0.083*** (0.025)	0.086*** (0.025)	0.029*** (0.007)	0.029*** (0.007)	0.026*** (0.008)	0.026*** (0.008)
Individual characteristics	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Economic characteristics		Yes	Yes	Yes		Yes	Yes	Yes
Labor characteristics			Yes	Yes			Yes	Yes
Land characteristics				Yes				Yes
Obs	21,130	21,130	17,689	17,326	21,130	21,130	17,689	17,326
Log likelihood	−14334.053	−14311.763	−12030.773	−11769.675	—	—	—	—
Wald chi2	139.87	174.42	198.11	212.52	—	—	—	—
R^2	—	—	—	—	0.071	0.073	0.072	0.072

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.

2. Changing the focus variable: use the land contract. In addition to the land certificate issued by the state, a land contract is also a guarantee for the daily implementation of agricultural land use rights (Huang & Ji, 2012). For this reason, this study uses land contracts to characterize land certification and replace the original land certificate. Table 4 reports the regression results of land contracts affecting the use of organic fertilizers. The results show that whether clustered to households or the community, the land contract positively increases the possibility of farmers using organic fertilizer at a significant level of 1%.

Table 4. Land certification and use of organic fertilizer: Robustness test II.

	Panel logit model
	(1)	(2)
Land contract	0.198*** (0.048)	0.198*** (0.045)
Control variables	Yes	Yes
Obs	19,925	19,925
Cluster unit	Household	community
Log likelihood	−13531.276	−13531.276
Wald chi2	225.73	225.73

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2.

3. Changing the dependent variable: use farm organic fertilizers and organic fertilizers. With the improvement of production and living conditions and the advancement of agricultural technology in China, traditional farm organic fertilizer represented by straw and manure has gradually been replaced by commercial organic fertilizer (Chu et al., 2012). However, it cannot be ignored that there is still a long historical process for commercial organic fertilizer to replace organic farm fertilizer completely. Therefore, this study uses the 2019 straw return variables, the 2017 and 2019 manure return variables, and the generalized organic fertilizer variables without considering their types from the database to estimate the empirical Equation (1)(1) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\alpha }_0+{\alpha }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\alpha }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$(1) again. Table 5 reports the regression results of the use of different types of organic fertilizers for land certification. Regardless of the types of organic fertilizer, land certification can significantly increase the probability of their use, which is consistent with existing research (Huang & Ji, 2012; M.S. Qian & Lu, 2020; Zhou & Wang, 2019). Moreover, the results in Column (6) show that, regardless of the type of organic fertilizer, farmers with the land certificate are nearly 30% more likely to use organic fertilizer than those without it. Compared with the previous certification effect that does not include the use of farm organic fertilizers (that is, commercial organic fertilizer), the incentives of land certification corresponding to the use of general organic fertilizers have almost doubled. These contributions are mainly from farm organic fertilizers, of which the straw is returned to the field (5.2%), and manure is returned to the field (13.9%).

Table 5. Land certification and use of organic fertilizer: Robustness test III.

	Returning straw to the field		Returning manure to the field		Organic fertilizer
	(1)	(2)	(3)	(4)	(5)	(6)
Land certification	0.054*** (0.010)	0.052*** (0.010)	0.113* (0.062)	0.139** (0.062)	0.222*** (0.043)	0.298*** (0.049)
Control variables		Yes		Yes		Yes
Obs	11,100	11,085	13,081	12,735	21,176	17,347
Log likelihood	−7471.614	−7382.299	−8189.529	−7900.022	−10408.621	−8473.438
Wald chi2	30.04	187.91	29.05	136.72	578.72	601.94

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2; All are logit models, but since only 2019 data is available for returning straw to the field, it is a cross-sectional logit model, and the others are panel logit models.

Further analysis

Generally, the behavior of farmers towards organic fertilizer use can be roughly divided into two categories; one is whether farmers use organic fertilizer, and the other is the expenditure and amount of organic fertilizers used by farmers. Therefore, after confirming that land certification significantly increases the probability of farmers’ organic fertilizer use. It is still necessary to further discuss the extent to which land certification will increase the expenditure and amount of organic fertilizers used by farmers.

Table 6 reports the regression results of land certification on farmers’ expenditure on organic fertilizers. Columns (1) and (2) use the panel ordinary least squares (OLS) model, and Columns (3) and (4) use the panel Tobit model. The estimation results of the OLS model show that land certification significantly increases farmers’ expenditure on organic fertilizers. Farmers who hold land certificates have an average of about 25 yuan more organic fertilizer per mu than those who do not. However, considering that the expenditure of organic fertilizer per mu of many farmers is 0 yuan and there are some extreme values simultaneously, the direct use of OLS estimation may be underestimated. Therefore, referring to the estimation method of L. Gao et al. (2011), the panel Tobit model is used to estimate this equation again. The estimation results of the Tobit model show that the OLS estimation method does indeed have the possibility of underestimation. Land certification increases the average farmer’s organic fertilizer expenditure by approximately 35 yuan per mu.

Table 6. Estimated results of land certification on farmers’ expenditure on organic fertilizers.

	OLS FE		Tobit FE
	(1)	(2)	(3)	(4)
Land certification	25.246** (9.899)	25.072** (10.083)	33.409*** (7.336)	35.738*** (7.350)
Control variables		Yes		Yes
Obs	15,409	14,985	15,409	14,985
Log likelihood	—	—	−25237.939	−24634.987
Wald chi2	—	—	20.74	142.32
R^2	0.082	0.083	—	—

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2; The panel Tobit model is merged to the left at 0 and merges to the right at 1000.

Table 7 reports the regression results of land certification on the amount of organic fertilizers used by farmers. Columns (1) and (2) show the results of the OLS method. Considering that there may be a self-selection problem for the pilot selection in the reform of land certification, Columns (3) and (4) use the Heckman two-step method. The results show that irrespective of the estimation method adopted, land certification significantly increases the amount of organic fertilizer used by farmers. Moreover, the estimated coefficients of land certification are very close. The average amount of organic fertilizer used by farmers with land certificates is about 23 kilograms per mu.

Table 7. Estimated results of land certification on the amount of organic fertilizers used by farmers.

	OLS		Heckman
	(1)	(2)	(3)	(4)
Land certification	24.011*** (7.907)	22.990*** (7.776)	24.029*** (8.001)	22.990*** (8.155)
Control variables		Yes		Yes
Obs	12,417	11,895	12,406	11,895
Log likelihood	—	—	−29627.069	−28841.005
Wald chi2	—	—	20.74	64.44
R^2	0.071	0.086	—	—

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2; Since the amount of organic fertilizer use is only available in 2019, the data is cross-sectional.

Mechanism of land certification on farmers’ organic fertilizer use

The above estimation results have proved the positive effect of land certification on the use of organic fertilizer. Furthermore, the issue that needs to be explored is how land certification affects farmers’ organic fertilizer use behavior. Combining the theoretical analysis in the previous section, this study begins with property rights expectations (whether land certification can bring benefits, yes = 1, no = 0) and the availability of agricultural subsidies (Whether to receive agricultural subsidies, yes = 1, no = 0). It draws on the method of Baron and Kenny (1986). Based on Equation (1), the following mediating effect model is further constructed to empirically test the mechanism of land certification affecting farmers’ use of organic fertilizer.

(2) $\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}={\beta }_0+{\beta }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\beta }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\omega }_{\mathrm{i}\mathrm{t}}$(2)

(3) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\gamma }_0+{\gamma }_1\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}+{\gamma }_2\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\gamma }_3{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\mu }_{\mathrm{i}\mathrm{t}}$(3)

The test idea of the mediating effect is as follows: on the basis of the significant positive coefficients of land certification in Column (5) of Table 2, we further estimate Equations (2)(2) $\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}={\beta }_0+{\beta }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\beta }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\omega }_{\mathrm{i}\mathrm{t}}$(2) and (3(3) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\gamma }_0+{\gamma }_1\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}+{\gamma }_2\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\gamma }_3{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\mu }_{\mathrm{i}\mathrm{t}}$(3) ), where Med is the two mediating variables concerned in this paper. That is, the expectation of farmer’s property rights and the availability of agricultural subsidies. If the coefficient β₁ in Equation (2)(2) $\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}={\beta }_0+{\beta }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\beta }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\omega }_{\mathrm{i}\mathrm{t}}$(2) and the coefficient γ₁ in Equation (3)(3) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\gamma }_0+{\gamma }_1\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}+{\gamma }_2\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\gamma }_3{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\mu }_{\mathrm{i}\mathrm{t}}$(3) are both significant, it means that land certification affects farmers’ organic fertilizer use behavior through the above-mentioned mediating variables. On this basis, if the γ₂ coefficient is still significantly positive, it means that the mediating variable plays a partial mediating role; if the γ₂ coefficient is not significant, it means that the mediating variable plays a complete mediating role. At the same time, if at least one of the coefficient β₁ in Equation (2)(2) $\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}={\beta }_0+{\beta }_1\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\beta }_2{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\omega }_{\mathrm{i}\mathrm{t}}$(2) and the coefficient γ₁ in Equation (3)(3) $\mathrm{O}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{\_}\mathrm{f}\mathrm{e}\mathrm{r}{\mathrm{t}}_{\mathrm{i}\mathrm{t}}={\gamma }_0+{\gamma }_1\mathrm{M}\mathrm{e}{\mathrm{d}}_{\mathrm{i}\mathrm{t}}+{\gamma }_2\mathrm{C}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}{\mathrm{n}}_{\mathrm{i}\mathrm{t}}+{\gamma }_3{\mathrm{X}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{i}}+{\lambda }_{\mathrm{t}}+{\mu }_{\mathrm{i}\mathrm{t}}$(3) is not significant, further Sobel test is needed for the interaction term β₁×γ₁. If the corresponding Z statistic is significant, it indicates that the mediating effect exists.

Table 8 reports the estimated results of the impact of land certification on the use of organic fertilizer by farmers. Judging from the estimated results of the first stage, the estimated coefficients of land certification on farmers’ property rights expectations and the availability of agricultural subsidies are 0.316 and 0.358, respectively, which comply with theoretical expectations, and both are significant at the 1% level. This shows that land certification can significantly improve the expectations of farmers’ property rights and increase the availability of agricultural subsidies (Zheng & Qian, 2022). According to the estimation results of the second stage, compared with the baseline regression results in Column (5) of Table 2, the estimated coefficient of land certification is still positive and significant at the 1% level, after considering the effects of farmers’ property expectation and agricultural subsidy availability. This indicates that the above mediating variables play a partial mediating role. In summary, it can be found that land certification mainly increases the possibility of farmers’ organic fertilizer use through channels such as stabilizing farmers’ property rights expectations and increasing the availability of agricultural subsidies.

Table 8. The effect mechanism of land certification on the use of organic fertilizer.

	Mediating effect model
	Channel I		Channel II
	Farmers’ property rights expectations	Organic fertilizer use	Availability of agricultural subsidies	Organic fertilizer use
Land certification	0.316*** (0.059)	0.153*** (0.043)	0.358*** (0.050)	0.149*** (0.044)
Farmers’ property rights expectations		0.136** (0.068)
Availability of agricultural subsidies				0.093*** (0.047)
Control variables	Yes	Yes	Yes	Yes
Obs	17,347	17,326	17,025	17,025
Log likelihood	−5569.546	−11764.025	−9699.515	−11555.291
Wald chi2	134.86	212.79	288.97	212.24

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2.

Heterogeneity analysis of land certification on farmers’ organic fertilizer use

The previous empirical evidence shows that land certification generally increases the probability of farmers using organic fertilizer. Furthermore, is there any heterogeneity in the impact of land certification on farmers’ organic fertilizer? To this end, this study tries to discuss heterogeneity from the perspectives of different production characteristics, property rights, and regional groupings.

Impact of land certification on farmers’ organic fertilizer use under different production characteristics

Existing documents have pointed out that land conservation behaviors of farmers with different land production scales vary. Therefore, this study draws on the practice of L. Qian et al. (2019). According to whether the land production area is larger than the sample mean, all farmers are divided into two types of small-scale and large-scale farmers. The estimated results in Table 9 show that land certification has a significant and positive impact on small-scale farmers, but has no effect on large-scale farmers’ use of organic fertilizers.

Table 9. Land certification and use of organic fertilizer: Heterogeneity test I.

	Different scales		Different crops
	Smaller scale	Larger scale	Food crops	Cash crops
Land certification	0.188*** (0.049)	0.003 (0.160)	0.153*** (0.056)	0.705 (0.730)
Observed difference	0.185***		−0.552**
Empirical p-value	0.000		0.043
Control variables	Yes	Yes	Yes	Yes
Obs	12,692	4,634	9,559	324
Log likelihood	−8636.002	−3114.467	−6409.619	−180.873
Wald chi2	178.76	93.11	84.95	3.16

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2.

At the same time, considering that different crops have different fertilizer requirements, the use of organic fertilizers may also be affected by the types of crops. Therefore, this study divides all farmers into two categories: farmers that grow food crops and farmers that grow cash crops. The estimation results show that land certification has a significant and positive impact on the use of organic fertilizer by farmers who grow food crops, but has no effect on the use of organic fertilizer by farmers who grow cash crops.

For the above results, the possible explanation is that whether it is a large-scale farmer or a farmer who grows cash crops, the main goal of their agricultural production is to maximize profits. Moreover, organic fertilizers may have slower effects and low yields compared to chemical fertilizers. Therefore, regardless of whether the land certification is completed, the organic fertilizer use behavior of these farmers will not change significantly in the short term. In contrast, a small-scale farmer or a farmer who grows food crops has a partially self-sufficient production orientation. To a large extent, cultivated land is still their means of production and living. Therefore, after land certification, they may cherish the cultivated land more and apply more organic fertilizer.

Impact of land certification on farmers’ organic fertilizer use under different property rights

Table 10 reports the regression results of land certification on using organic fertilizers under different property rights. The estimation results show that, for different plots, land certification has a positive and significant impact on the use of organic fertilizers on contracted plots, but has no effect on the use of organic fertilizers on transferred plots. This is in line with the research of Zhou and Wang (2019). The possible reason is that the new round of land certification in China is mainly aimed at the reform of land property rights of contracted farmers. It can enhance the tenure security of contracted plots but has little impact on transferred plots.

Table 10. Land certification and use of organic fertilizer: Heterogeneity test II.

	Different plots		Land adjustment experience
	Contracted land	Transferred land	Yes	No
Land certification	0.139*** (0.045)	0.183 (0.130)	0.097 (0.072)	0.251*** (0.058)
Observed difference	−0.044***		−0.154***
Empirical p-value	0.000		0.000
Control variables	Yes	Yes	Yes	Yes
Obs	14,934	2,392	6,324	9,344
Log likelihood	−10171.284	−1596.608	−4292.760	−6361.565
Wald chi2	173.90	42.05	88.86	122.37

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2.

Furthermore, under China’s household contracting system, farmers’ contracted land is often reallocated along with population changes, which triggers the problem of land adjustment and undermines the stability of property rights. To this end, this study draws on the method of M.S. Qian and Lu (2020), based on whether farmers have experienced land adjustment since the second round of contracts, and divides all farmers into those with land adjustment experience and those with no land adjustment experience. The estimation results show that land certification mainly has a positive and significant impact on farmers with no land adjustment experience, but has no effect on those with land adjustment experience.

For the above results, the possible explanation is that for farmers with land adjustment experience, may lead them to have a distrustful attitude towards land certification. Since land adjustment has occurred in the past, regardless of whether the land has been certified, it is still more likely to happen in the future. On the contrary, for farmers without land adjustment experience, their land has been stable, and the property rights-oriented land certification will only strengthen rather than weaken the stability of land property rights. Therefore, compared with those with land adjustment experience, farmers without land adjustment experience are more willing to make long-term investments in land and increase the probability of organic fertilizer use.

Impact of land certification on farmers’ organic fertilizer use under different regional groupings

Table 11 reports the regression results of land certification on the use of organic fertilizer under different regional groupings. China has a vast territory with different natural conditions and economic development levels, and the impact of land certification on farmers’ use of organic fertilizers in different regions may not be exactly the same. In terms of geographical divisions, due to the relatively backward economy in the western region, most farmers rely on agriculture as their industry and land for their livelihoods. According to the database statistics, the average annual agricultural net income of farmers in the western region was only 3,264 yuan, which was lower than 4,193 yuan in the central region and 7,366 yuan in the eastern region. Therefore, compared with farmers in the eastern and central regions, the impact of land certification on those who live in the western region may be more direct, and they are more likely to impose long-term investments such as organic fertilizers. The results in Table 11 show that, as expected, land certification mainly has a positive and significant impact on farmers’ use of organic fertilizer in the western region areas.

Table 11. Land certification and use of organic fertilizer: Heterogeneity test III.

	Geographical regions
	(1) East	(2) Middle	(3) West
Land certification	−0.034 (0.068)	0.114 (0.076)	0.184*** (0.071)
Control variables	Yes	Yes	Yes
Obs	7,108	4,792	5,426
Log likelihood	−4758.977	−3242.801	−3481.968
Wald chi2	84.79	67.90	56.27

Standard errors are in parentheses; ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; The control variables are the same as those in column (5) of Table 2; Geographical divisions refer to the standards of the National Bureau of Statistics of China (http://www.stats.gov.cn/ztjc/zthd/sjtjr/dejtjkfr/tjkp/201106/t20110613_71947.htm).

Conclusion and discussion

Conclusion

Deepening the reform of the land tenure system can theoretically improve the quality of cultivated land and thus lay a material foundation for food security. Based on nationally representative panel data and multiple panel models, this paper explores the relationship between land tenure system reform and cultivated land quality by examining the impact of China’s new round of land certification on farmers’ use of organic fertilizers. We provide robust empirical evidence that land certification significantly increases the probability of farmers using organic fertilizer, as well as the expenditure and amount of organic fertilizers used by farmers, which contribute to the improvement of cultivated land quality. The results of the mechanism test indicate that land certification helps stabilize farmers’ expectations of property rights, improve the availability of agricultural subsidies, and thereby increase farmers’ use of organic fertilizers.

The results of the heterogeneity analysis show that compared with large-scale farmers and farmers planting cash crops, land certification mainly has a positive and significant effect on small-scale farmers or farmers who grow food crops. Meanwhile, compared with farmers with transferred land and farmers with land adjustment experience, land certification mainly has a positive and significant effect on farmers with contracted land or without land adjustment experience. Moreover, compared with the eastern and central regions, land certification mainly has a positive and significant effect on the western region.

Discussion

Currently, with the spread of COVID-19, the ongoing Russia-Ukraine conflict, and frequent extreme weather disasters, food security based on the quality of cultivated land is facing more severe challenges than ever before. Therefore, the above research findings may have significant policy implications for China and other developing countries.

First, it is necessary for developing countries to carry out large-scale reform of land certification and thus establish the property rights system foundation of farmland quality protection. Insecure land tenure is a common challenge faced by many developing countries, which weakens farmers’ enthusiasm for production to some extent. People with constant property have perseverance. Issuing land certificates to farmers can effectively stabilize farmers’ expectations of property rights and improve their availability of agricultural subsidies. When farmers’ land rights and interests are protected by law, they are more likely to have the confidence to make long-term investments in cultivated land.

Second, the government should guide large-scale farmers and farmers who grow cash crops to strengthen the protection of cultivated land and promote the sustainable use of cultivated land. Large-scale farmers and farmers who grow cash crops usually prioritize profit maximization and carry out market-oriented agricultural production. They may focus more on short-term gains than small-scale farmers and farmers who grow food crops. Therefore, in the process of agricultural production, they should be guided to establish the awareness of farmland protection, and insist on attaching equal importance to short-term interests and long-term interests.

Third, the government should equally protect the land rights and interests of farmers who transfer their land and those who have undergone land adjustment and improve the security of their land property rights. Insecure land property rights can easily lead farmers to realize that cultivated land is not their own, resulting in the irrational use of cultivated land. Regardless of the status of land property rights in the past, the most important thing is to protect land property rights equally going forward. Only in this way will farmers cherish the cultivated land more and apply more organic fertilizers to it.

Finally, developed areas need to increase transfer payments to agriculture and raise the income level of local farmers. Compared with less developed areas, there are more off-farm opportunities in developed areas, and the opportunity cost for farmers to farm is also higher, so local farmers may pay less attention to the quality of cultivated land. At such time, the developed area must use its economic advantages to support agricultural development. For example, developed areas can allocate special funds for the research and development of organic fertilizers with low cost and high yield or provide corresponding subsidies to farmers who buy organic fertilizers.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was supported by the Major Program of National Fund of Philosophy and Social Science of China (19ZDA088), the National Natural Science Foundation of China (72273061),the Zhejiang New Think Tank Research Project (22ZK23YB) and the Fundamental Research Funds for the Central Universities (20221121).