CHAPTER 5. Analysis of Farmers’ Responses to
3. Analysis of Farmers’ Responses
3.1. Analysis Model for Farmers’ Responses
In order to determine the extent of farmers’ acceptance of sustainable agriculture, the responses to the survey questionnaire for farmers was analyzed using Heckman model. The Heckman model estimates the parameters for the sample selection equation, based on the maximum likelihood method. In this process, Inverse Mill’s Ratio (IMR, ), is calculated and in order to consistently estimate parameters, is included as an independent variable in 2-step resulting equation. IMR is calculated as shown in <Eq. 5-1>.
(5-1)
where, φrepresents a probability density function of standard normal distribution, and Ф a cumulative distribution function In Heckman’s 2-step estimation procedure, the sample selection equation is expressed as shown in
<Eq. 5-2>.
∗
0if ∗ 0; ~ 0,1 (5-2)
1if ∗ 0
where, ∗ represents the farmers’ utility level unobserved. If the utility level ∗ is bigger than 0, it means that the farmer is inclined (zi) to accept sustainable agriculture. Heckman’s2-step resulting equation is expressed as shown in <Eq.
5-3>.
∗
∗if 1 (5-3)
if 0
where, yi represents the extent of producers’ acceptance of sustainable agriculture, xi a vector of the independent variables that affect yi, and β a vector of the coefficient to be estimated.
Explanatory variables used in the analysis model and average values of these variables are shown in <Table 5-26>. First of all, as social and economic characteristics of individual farmers are likely to affect their inclination to accept sustainable agriculture, area of their farmland (ACR), age (AGE), level of education (EDU), main item - flowers (FLOWERS), mountainous cultivation areas (MOUNTAIN), experiences in farming (CAREER), income (INC), main item - fruits (FRUIT) and region (REGION) were included as explanatory variables. Here, for the farmers whose main item was fruits, the variable value was set to 1 and for the farmers who grew other items, it was set to 0. With regard to the variable of REGION, 1 was set for the farmers in Jeollanam-do region and 0 for those in other regions.
Next, in order to examine the correlation among the extent of farmers practicing sustainable agriculture, how well they recognized it, and how much they were inclined to practice it, the following were added as explanatory variables: farmers’ recognition of sustainable agriculture (RECOG), extent of recognition of the importance of adopting it (IMP), and extent of farmers’
practicing it (STATUS). At this time, the farmers who recognized sustainable agriculture was given the variable value of 1, and the other farmers 0.
Lastly, to determine the correlation among farmers’ inclination to adopt sustainable agriculture and their recognition of its environmental, economic and social aspects, the following explanatory variables were added: experiences of conflicts with neighbor (CONFLICT); income difference (INCDIFF), economic issue (ECOPRO), environmental and social problems (ENVSOCPRO), and co-works with neighbors (COWORK). Here, for CONFLICT variable, the farmers who answered that they experienced conflict with neighbors while practicing sustainable agriculture were given the value of 1, and the others 0. For ECOPRO, the farmers who recognized that economic problems are the biggest problems the Korean agriculture was facing were given 1, and the others 0. ENVSOCPRO was set the same way <Table 5-26>.
Table 5-26. Explanation and technical statistics of variables used in the analysis
Variable Description Average Standard
Deviation No. of Samples ACR Cultivation area (3.3m2) 8667.6 10568.5 291
AGE Farmer age (years) 55.8 9.2 284
EDU Level of education (up to middle school
=1,..graduate school or higher =5) 3.0 1.2 291 FLOWERS Flowers as main item (Flowers =1,
Others=0) 0.0 0.2 291
MOUNTAIN Zone (Semi-mountainous and
mountainous region =1, Others=0) 0.4 0.5 246 CAREER Farmer’s career in experience (years) 17.4 13.1 289 INC Change in farmer’s income (Likert scale
of 5 points) 2.6 1.5 291
FRUIT Fruits as main item (Fruits =1, Others=0) 0.3 0.5 291 REGION Region (Jeollanam-do=1, Others=0) 0.1 0.4 291 RECOG Recognition of sustainable
agriculture(Recognized=1, Others=0) 0.8 0.6 291 IMP Importance of adopting sustainable
agriculture(Likert scale of 5 points) 4.4 0.8 291
STATUS
Status of practicing sustainable agriculture (conventional farmer=1, farmer practicing sustainable agriculture=2, farmer in transition =3)
2.2 0.7 289
CONFLICT
Experience of conflict with neighbors while practicing sustainable
agriculture(experienced conflicts =1, Not experienced conflicts =0)
0.4 0.5 205
INCDIFF Agreement to alleviating income difference in sustainable agriculture (Likert scale of 5 points)
3.2 1.1 287
ECOPRO
Recognition of problems the Korean agriculture is facing
(Economic problem =1, others=0) 0.7 0.5 291
ENVSOCPRO
Recognition of problems the Korean agriculture is facing
(Environmental/social problem =1, others=0)
0.2 0.4 291
COWORK Recognition of benefits of co-work in sustainable agriculture(Likert scale of 5 points)
3.4 1.1 290
3.2. Results of Analysis of Farmers’ Responses
According to the result of estimation by Probit model in Step 1, INC, IMP and CONFLICT had positive signs, being identified as significant variables that affect farmers to select sustainable agriculture. In other words, it was more likely for farmers with higher income, those who recognized that it was importance for Korea for them to adopt sustainable agriculture, and those who had experienced conflicts with neighbors while practicing sustainable agriculture, to select sustainable agriculture. Though the statistical significance was low, farmers with larger cultivation area (ACR), shorter career in farming (CAREER), or producing fruits as their main items (FRUIT) were less inclined to adopt sustainable agriculture, compared to the others. Also, farmers who were farming in Jeollanam-do provision (REGION) and those who had recognized sustainable agriculture (RECOG) were shown to be more inclined to adopt sustainable agriculture<Table 5-27>.
Next in Step 2, the impacts of the individuals’ socio-economic characteristics on their inclination to practice sustainable agriculture were regression-analyzed using Weighted Least Square (WLS). FLOWERS was significantly estimated to have a negative sign, while STATUS and ENVSOC-PRO were estimated to have a positive sign. In other words, farmers whose main item was flowers benefited less from practicing sustainable agriculture, compared to those whose main item was not flowers, and thus they were less inclined to practice sustainable agriculture. Also, farmers who were currently practicing sustainable agriculture, along with the more they recognized that Korean agriculture had serious environmental and social problems, they more they were inclined to practice sustainable agriculture.
Though statistical significance was low, INCDIFF,COWORK and ECOPRO were found to have positive signs. In other words, the more farmers recognized sustainable agriculture as a practice that could alleviate income difference and help promote unity or fellowship of the community through co-work, and the more they recognized that economic problems were the most serious problems Korean agriculture was facing, the greater was their inclination to practice sustainable agriculture.
Table 5-27. Results of regression analysis of famers’ responses (Heckman’s 2-tep estimation method)
Answer
Step 1(Probit) (Dependent variable: sustainable
agriculture or not)
Step 2(WLS)
(Dependent variable: percentage of practicing sustainable agriculture) Coefficient
Value Standard Error Coefficient
Value Standard Error
ONE -1.456** 0.651 1.845* 0.997
ACR -0.000 0.000 - -
AGE 0.001 0.001 0.020 0.012
EDU - - 0.073 0.082
FLOWERS - - -1.143*** 0.407
MOUNTAIN - - 0.00025 0.000
CAREER -0.004 0.008 - -
INC 0.132* 0.074 0.094 0.071
FRUIT -0.299 0.214 - -
REGION 0.307 0.301 - -
RECOG 0.126 0.166 - -
IMP 0.557*** 0.126 - -
STATUS - - 0.002* 0.001
CONFLICT 0.000** 0.000 - -
INCDIFF 0.001 0.001 0.004 0.096
ECOPRO - - 0.430 0.338
ENVSOCPR - - 0.923** 0.386
COWORK - - 0.001 0.001
LAMBDA - - -1.446* 0.839
Chi-square Number of obs.
Selected obs.
44.07(0.0000) 291 176
Note: 1) ***, ** and * are significant at the levels of 1%, 5% and 10% respectively.