The influencing factors and income effects of green prevention-control technology adoption — An empirical analysis based on the survey data of 792 vegetable growers
-
摘要: 本文以山东省寿光市蔬菜种植户为例, 对菜农采纳绿色防控技术的影响因素和收入效应进行定量分析, 以便为绿色防控技术在蔬菜种植行业的有效推广、促进菜农绿色生产和提质增收提供政策参考。基于山东寿光792户菜农的微观数据, 采用赫克曼矫正法模型从“采纳决策”和“采纳程度”两个方面探究了影响菜农绿色防控技术采纳的影响因素, 运用内生转换回归模型分析菜农绿色防控技术采纳对其收入的平均处理效应。研究发现: 1)菜农对绿色防控技术的采纳程度不足, 样本农户中只采纳1项和2项绿色防控技术的数量最多, 占总样本比重分别为36.87%和27.27%; 2)菜农绿色防控技术认知水平、接受质检和培训经历、会主动利用互联网搜集信息对菜农的绿色防控技术采纳行为显著正相关, 地块数量、信息设备数量对菜农的绿色防控技术采纳行为显著负相关; 3)人均收入超过样本平均水平的菜农更乐于接受绿色防控技术, 且该技术的采用可增加菜农年均收入的比重为7.2%。据此, 应发挥绿色防控技术的增收效应, 激发菜农采纳新技术的内生动力, 提出政府应加快健全绿色防控技术推广机制, 完善风险补偿机制、给予菜农技术政策补贴, 增强菜农的科学技术意识、积极推动农户适度规模经营等政策建议。Abstract: To comply with the agricultural production trend of protecting the ecological environment, reducing the application of chemical pesticides, comprehensively developing green agriculture, and helping the rural areas to achieve long-term development, the Chinese government made efforts to promote the application of green prevention-control technology; however, the application level of this technology in China is not high and existing research on the adoption of green prevention-control technology as a method selection by farmers is insufficient. To promote the adoption of green prevention-control technologies by vegetable farmers, achieve green production and quality improvement and income increase, provide policy reference for the effective promotion of green prevention-control technologies in the vegetable planting industry, and enrich the promotion theory of green prevention-control technologies, this study used vegetable growers in Shouguang City, Shandong Province as an example to quantitatively analyze the influencing factors and income effects of vegetable farmers’ adoption of green prevention-control technologies. Based on the micro-data of 792 vegetable farmers, this study characterized the behavior of vegetable farmers in adopting green prevention-control technology according to two aspects, adoption decision and adoption degree, explored the influencing factors affecting the adoption of green prevention-control technology of vegetable farmers using the Heckman correction method model. And the average treatment effect of the adoption of green prevention-control technology of vegetable farmers on their income was analyzed by using an endogenous conversion regression model. The study found that the degree of adoption of green prevention-control technologies by vegetable farmers was insufficient. The numbers of vegetable farmers adopting one and two green prevention-control technologies were the largest, which were 292 households and 216 households, respectively, accounting for 36.87% and 27.27% of the total sample, respectively. The level of awareness of green prevention-control technology of vegetable farmers, the experience of getting quality testing and training, and the active use of the internet in collection of information were significantly positively correlated with the adoption of green prevention-control technologies by vegetable farmers. Factors such as the number of plots owned, the number of information devices significantly negatively correlated with the adoption of green prevention and control technologies among vegetable farmers. Vegetable farmers whose per capita income exceeded the average sample level were more willing to accept green prevention-control technology, and the use of this technology could increase the proportion of the average annual income of vegetable farmers by 7.2%. Therefore, a variety of factors, such as family, information, and government policies, affect the decision-making and adoption degree of green prevention-control technologies by vegetable farmers; moreover, the adoption of green prevention-control technologies has a positive impact on the income of vegetable farmers. Based on this, we should give full play to the increase in the income effect of green prevention-control technology, stimulate the endogenous motivation of vegetable farmers to adopt new technologies, and propose that the government should accelerate the improvement of the promotion mechanism of green prevention-control technology, improve the risk compensation mechanism, give vegetable farmers science and technology policy subsidies, enhance the scientific and technological awareness of vegetable farmers, and actively promote the moderate-scale operation of farmers.
-
表 1 样本菜农采纳不同绿色防控技术的分布情况
Table 1. Distribution of different green prevention-control technologies adopted by sample vegetable famers
农药防治
Pesticide control物理防治
Physical control生物防治
Biotic-control生态调控
Ecological regulation采纳户数
Number of adopted households567 332 267 136 未采纳户数
Number of not
adopted household224 459 524 655 总户数
Total number of households792 792 792 792 表 2 菜农绿色防控技术采纳影响因素模型的变量定义、预期方向及描述性统计
Table 2. Variable definition, expected direction and descriptive statistics of influencing factors model of vegetable farmers’ green prevention-control technology adoption
变量名称
Variable name变量含义
Variable meaning预期方向
Expected direction均值
Mean标准差
Standard deviation因变量
Dependent variable是否使用绿色防控技术
Whether to use green prevention-
control technology1=是, 0=否
1=yes, 0=no0.88 0.32 采纳绿色防控技术个数
Number of green prevention-control
technologies adopted农户采纳该技术的实际个数
Actual number of farmers adopting
this technology1.65 1.01 自变量
Independent variable个人因素
Personal element性别
Gender1=男性, 0=女性
1=male, 0=female+ 0.60 0.49 年龄
Age农户的年龄
Farmer’s age− 51.41 8.83 年龄平方
Age square农户年龄的平方
Square of farmer’s age− 2720.84 914.55 教育
Education农户的上学时长(年)
Schooling years of farmers+ 7.79 2.96 家庭因素
Family factors种植人数
Number of planters农户家庭主营蔬菜种植的人数
Number of farmers mainly engaged in
vegetable cultivation− 2.06 0.62 蔬菜面积
Vegetable area农户目前种植蔬菜的实际面积
Actual area of vegetables at present (hm2)+ 0.27 0.35 地块数
Number of plots农户种植的实际地块数
Actual number of plots planted by farmers− 2.04 1.63 信息因素
Information factors互联网
Internet1=是, 0=否
1=yes, 0=no+ 0.48 0.50 信息设备
Information devices电脑、智能手机的数量
Number of computers and smart phones+ 2.31 0.95 您是否会利用信息设备主动搜集信息
Will you actively collect information by using information equipment?1=是, 0=否
1=yes, 0=no+ 0.86 0.35 政府因素
Government factors种植技术
Planting technique接受技术培训次数
Number of technical trainings received+ 3.25 1.01 质量检测
Quality detection接受质量检测的次数
Number of times of quality inspection+ 0.95 0.22 安全培训
Safety training接受安全培训次数
Number of safety trainings+ 0.36 0.48 指导化肥
Guide fertilizer接受指导化肥次数
Number of times of receiving fertilizer instruction+ 0.76 0.43 识别变量
Identification variable您觉得绿色防控技术有效吗
Do you think that green prevention and control
technology is effective?1=是, 0=否
1=yes, 0=no+ 0.80 0.40 表 3 绿色防控技术采纳的收入效应模型变量定义、预期方向及描述性统计
Table 3. Variable definition, expected direction and descriptive statistics of revenue effect model of adoptation of green prevention-control technologies
变量名称
Variable name变量含义
Variable meaning预期方向
Expected direction均值
Mean标准差
Standard deviation因变量
Dependent variable农业收入
Farming income每公顷净收益对数
Logarithm of average net income per hectare (¥)13.18 0.79 自变量
Independent variable是否使用绿色防控技术
Whether to use green prevention
and control technology1=是, 0=否
1=yes, 0=no+ 0.88 0.32 采纳绿色防控技术个数
Number of green prevention and
control technologies adopted农户采纳该技术的实际个数
Actual number of farmers
adopting this technology+ 1.65 1.01 表 4 菜农绿色防控技术采纳决策和采纳程度影响因素的估计结果及OLS回归和Logit回归对比
Table 4. Estimation of factors influencing the adoption decision and adoption degree of green prevention-control technology of vegetable farmers, and the comparison of OLS regression and Logit regression
变量
Variable赫克曼矫正法
Heckman correction methodOls回归
Ols regressionLogit回归
Logit regression采纳程度
Adoptation degree采纳决策
Adoptation decision采纳程度
Adoptation degree采纳行为
Adoptation behavior系数
Coefficient标准误
Standard error系数
Coefficient标准误
Standard error系数
Coefficient标准误
Standard error系数
Coefficient标准误
Standard error性别 Gender 0.050 0.070 0.023 0.138 0.068 0.075 0.087 0.255 年龄 Age 0.002* 0.033 0.071 0.053 0.057* 0.032 0.122 0.095 年龄平方 Age square 0.000 0.000 −0.001 0.001 −0.001* 0.000 −0.001 0.001 教育 Education 0.016* 0.013 0.027 0.024 0.027** 0.013 0.061 0.044 种植人数
Number of planters−0.013 0.055 −0.124 0.098 −0.090 0.056 −0.237 0.176 蔬菜面积 Vegetable area 0.008 0.009 −0.022 0.014 −0.009 0.008 −0.039* 0.024 地块数 Number of plots −0.058** 0.030 0.163** 0.071 0.012 0.027 0.309** 0.141 网络 Internet 0.125 0.087 0.301** 0.140 0.314*** 0.075 0.571** 0.267 信息设备
Information devices−0.386*** 0.112 0.218 0.186 −0.161 0.112 0.413 0.341 搜集信息
Gather information0.016 0.046 0.146 0.090 0.098** 0.045 0.250 0.167 种植技术 Planting technique −0.041 0.039 −0.111 0.073 −0.109*** 0.038 −0.228* 0.136 质量检测 Quality detection 0.474** 0.197 0.523** 0.233 0.078 0.158 0.866** 0.412 安全培训 Safety training 0.126 0.131 0.813*** 0.175 0.328*** 0.077 1.591*** 0.375 指导化肥 Guide fertilizer 0.121 0.083 −0.059 0.147 0.074 0.087 −0.119 0.269 有效性认知
Effective cognition0.262* 0.152 0.472* 0.279 常数项 Constant term 3.022*** 1.033 −1.359 1.529 0.487 0.895 −2.096 2.748 逆米尔斯比率 Mills lambda −1.414** ***、**、* 分别表示变量在P<0.01、P<0.05和P<0.1置信水平显著。***, ** and * respectively indicate that the variables are significant at P<0.01, P<0.05 and P<0.1 confidence levels. 表 5 绿色防控技术采纳行为对采纳和未采纳农户农业收入的影响对比
Table 5. Comparison of the impact of green prevention-control technology adoption behvior on the agricultural income of farmers adopting and unaccepting technology
变量
Variable绿色防控技术采纳模型
Green prevention-control technology adoption model农业收入 Farming income 采纳技术
Adopting technology未采纳技术
Unaccepting technology系数
Coefficient标准差
Standard deviation系数
Coefficient标准差
Standard
deviation系数
Coefficient标准差
Standard
deviation性别 Gender −0.227 0.167 −0.790 0.318 −0.060 0.825 年龄 Age 0.005 0.009 0.770 0.146 0.300 0.361 年龄平方 Age square −0.051** 0.001 2.060** 0.001 1.830* 0.004 教育 Education −0.036 0.027 3.610*** 0.059 8.400*** 0.145 种植人数 Number of planters 0.186 0.099 −1.900* 0.266 3.320*** 0.622 蔬菜面积 Vegetable area 0.013 0.016 −1.800* 0.039 2.100** 0.094 地块数 Number of plots 0.421 0.141 −1.140 0.145 0.04 0.295 网络 Internet 0.034 0.139 0.370*** 0.386 1.670* 0.822 信息设备 Information devices −0.317 0.195 −0.516** 0.481 0.526 1.254 搜集信息 Gather information 0.484*** 0.087 −2.050** 0.254 −0.250 0.488 种植技术 Planting technique 0.028 0.074 0.080** 0.213 0.544*** 0.417 质量检测 Quality detection 0.072 0.245 −0.820 0.562 −0.430 1.849 安全培训 Safety training −0.070 0.154 1.480 0.682 6.040*** 0.828 指导化肥 Guide fertilizer 0.075 0.148 −2.260** 0.359 0.750 0.977 有效性认知 Effective cognition 0.116 0.149 — — — — 常数项 Constant term −0.420 1.594 1.294 0.430 2.062 1.934 rho1 — — −0.286** 0.176 — — rho2 — — — — 0.787*** 0.435 Log likelihood −256.583 ***、**、* 分别表示变量在P<0.01、P<0.05和P<0.1置信水平显著。***, ** and * respectively indicate that the variables are significant at P<0.01, P<0.05 and P<0.1 confidence levels. 表 6 菜农采纳绿色防控技术对农业收入影响的平均处理效应
Table 6. Average treatment effect of the adoption of green prevention-control technology on agricultural income of vegetable farmers
结果均值 Result mean 平均处理效应(ATT)
Average treatment effectt值
t value变化率
Rate of change采纳
Adopting technology未采纳
Unaccepting technology农业收入
Farming income13.479 12.515 0.964 54.247*** 0.072 -
[1] 王佳新, 李媛, 王秀东, 等. 中国农药使用现状及展望[J]. 农业展望, 2017, 13(2): 56−60 doi: 10.3969/j.issn.1673-3908.2017.02.012WANG J X, LI Y, WANG X D, et al. Status quo of pesticide use in China and its outlook[J]. Agricultural Outlook, 2017, 13(2): 56−60 doi: 10.3969/j.issn.1673-3908.2017.02.012 [2] 杨程方, 郑少锋, 杨宁. 信息素养、绿色防控技术采用行为对农户收入的影响[J]. 中国生态农业学报(中英文), 2020, 28(11): 1823−1834YANG C F, ZHENG S F, YANG N. The impact of information literacy and green prevention-control technology adoption behavior on farmer household income[J]. Chinese Journal of Eco-Agriculture, 2020, 28(11): 1823−1834 [3] 崔晓, 张屹山. 中国农业环境效率与环境全要素生产率分析[J]. 中国农村经济, 2014(8): 4−16CUI X, ZHANG Y S. Analysis of agricultural environmental efficiency and environmental total factor productivity in China[J]. Chinese Rural Economy, 2014(8): 4−16 [4] ECOBICHON D J. Pesticide use in developing countries[J]. Toxicology, 2001, 160(1/2/3): 27−33 [5] 刘晓雪. 新时代乡村振兴战略的新要求−2018年中央一号文件解读[J]. 毛泽东邓小平理论研究, 2018(3): 13−20, 107LIU X X. The new requirements for the strategy of rural revitalization in a new era: an interpretation of the “Central Document No. 1” in 2018[J]. Studies on Mao Zedong and Deng Xiaoping Theories, 2018(3): 13−20, 107 [6] 高杨, 牛子恒. 风险厌恶、信息获取能力与农户绿色防控技术采纳行为分析[J]. 中国农村经济, 2019(8): 109−127GAO Y, NIU Z H. Risk aversion, information acquisition ability and farmers’ adoption behavior of green control techniques[J]. Chinese Rural Economy, 2019(8): 109−127 [7] 李太平, 张锋, 胡浩. 中国化肥面源污染EKC验证及其驱动因素[J]. 中国人口·资源与环境, 2011, 21(11): 118−123 doi: 10.3969/j.issn.1002-2104.2011.11.020LI T P, ZHANG F, HU H. Authentication of the Kuznets curve in agriculture non-point source pollution and its drivers analysis[J]. China Population, Resources and Environment, 2011, 21(11): 118−123 doi: 10.3969/j.issn.1002-2104.2011.11.020 [8] 杨普云. 农作物有害生物全程绿色防控技术模式的集成与推广应用[J]. 中国植保导刊, 2018, 38(4): 21−25 doi: 10.3969/j.issn.1672-6820.2018.04.004YANG P Y. Integration and application of the whole-process technical scheme of green management against crop pests[J]. China Plant Protection, 2018, 38(4): 21−25 doi: 10.3969/j.issn.1672-6820.2018.04.004 [9] 李培林. 理性选择理论面临的挑战及其出路[J]. 社会学研究, 2001, 16(6): 43−55LI P L. Challenge to the rational choice theory (RCT) and the way out[J]. Sociological Research, 2001, 16(6): 43−55 [10] 陈会英, 郑强国. 中国农户科技水平影响因素与对策研究[J]. 农业技术经济, 2001(2): 21−26 doi: 10.3969/j.issn.1000-6370.2001.02.006CHEN H Y, ZHENG Q G. Research on influencing factors and countermeasures of Chinese farmers’ science and technology level[J]. Journal of Agrotechnical Economics, 2001(2): 21−26 doi: 10.3969/j.issn.1000-6370.2001.02.006 [11] 乔娟, 曹蕾. 基于食品质量安全的农户认知、行为、态度和意愿分析−以江苏省如东县长沙镇水产养殖为例[J]. 中国畜牧杂志, 2009, 45(14): 24−28QIAO J, CAO L. Analyses of farmers’ awareness, behavior, attitude and willingness based on food quality safety — Take Changsha Town, Rudong County of Jiangsu Province as an example[J]. Chinese Journal of Animal Science, 2009, 45(14): 24−28 [12] 杨玉苹, 孙炜琳, 朱立志. 农户生物菌肥购买意愿及行为的影响因素研究−基于山东省设施菜农的调研数据[J]. 中国农业资源与区划, 2019, 40(2): 49−55, 78YANG Y P, SUN W L, ZHU L Z. The influence factors of farmer’s purchasing intension and behavior of biological fertilizer in Shandong[J]. Chinese Journal of Agricultural Resources and Regional Planning, 2019, 40(2): 49−55, 78 [13] DOSS C R, MORRIS M L. How does gender affect the adoption of agricultural innovations?[J]. Agricultural Economics, 2000, 25(1): 27−39 doi: 10.1111/j.1574-0862.2001.tb00233.x [14] DOSS C R. Designing agricultural technology for African women farmers: lessons from 25 years of experience[J]. World Development, 2001, 29(12): 2075−2092 doi: 10.1016/S0305-750X(01)00088-2 [15] 赵建欣, 张忠根. 对农户种植安全蔬菜的影响因素分析−基于对山东、河北两省菜农的调查[J]. 国际商务(对外经济贸易大学学报), 2008(2): 52−57ZHAO J X, ZHANG Z G. The determinants of farmer’ vegetable production: based on the survey of Hebei and Shandong Provinces[J]. International Business, 2008(2): 52−57 [16] 张云华, 马九杰, 孔祥智, 等. 农户采用无公害和绿色农药行为的影响因素分析−对山西、陕西和山东15县(市)的实证分析[J]. 中国农村经济, 2004(1): 41−49ZHANG Y H, MA J J, KONG X Z, et al. The behaviours of farmers’ adopting pollution-free and green pesticides and analysis of the causal factors — A positive analysis based on surveys in 15 counties of Shanxi, Shaanxi and Shandong Provinces[J]. Chinese Rural Economy, 2004(1): 41−49 [17] 张复宏, 宋晓丽, 霍明. 果农对过量施肥的认知与测土配方施肥技术采纳行为的影响因素分析−基于山东省9个县(区、市)苹果种植户的调查[J]. 中国农村观察, 2017(3): 117−130ZHANG F H, SONG X L, HUO M. Excess fertilizer application and growers’ adoption behavior for soil testing for fertilizer formulation and their determinants: an empirical analysis based on survey data from apple growers in 9 counties of Shandong Province[J]. China Rural Survey, 2017(3): 117−130 [18] 程杰贤, 郑少锋. 政府规制对农户生产行为的影响−基于区域品牌农产品质量安全视角[J]. 西北农林科技大学学报(社会科学版), 2018, 18(2): 115−122CHENG J X, ZHENG S F. Analysis of influence of government regulation on peasant households’ production behavior based on regional brand agricultural product quality and safety[J]. Journal of Northwest A & F University (Social Science Edition), 2018, 18(2): 115−122 [19] 黄晓慧, 王礼力, 陆迁. 农户认知、政府支持与农户水土保持技术采用行为研究−基于黄土高原1152户农户的调查研究[J]. 干旱区资源与环境, 2019, 33(3): 21−25HUANG X H, WANG L L, LU Q. Farmers’ cognition, government support and farmers’ soil and water conservation technology adoption in Loess Plateau[J]. Journal of Arid Land Resources and Environment, 2019, 33(3): 21−25 [20] YODER F R, WILLARD W, COCHRANE. Farm prices: myth and reality. pp. vii, 189. Min Neapolis: University of Minnesota Press, 1958. $4.00[J]. The ANNALS of the American Academy of Political and Social Science, 1959, 324(1): 174−175 doi: 10.1177/000271625932400155 [21] 黄祖辉, 钱峰燕. 技术进步对我国农民收入的影响及对策分析[J]. 中国农村经济, 2003(12): 11−17HUANG Z H, QIAN F Y. Analysis of the impacts of agricultural technological progress on farmers’ incomes in China with some policy implications[J]. Chinese Rural Economy, 2003(12): 11−17 [22] MERCY A, ANEANI F, OFORI S, et al. Analysis of farmers adoption behaviour of CRIG recommended technologies as a package: the case of some self help cocoa farmer associations in the eastern region of Ghana[J]. Agricultural Sciences, 2015, 6(6): 601−608 doi: 10.4236/as.2015.66059 [23] 张宽, 邓鑫, 沈倩岭, 等. 农业技术进步、农村劳动力转移与农民收入−基于农业劳动生产率的分组PVAR模型分析[J]. 农业技术经济, 2017(6): 28−41ZHANG K, DENG X, SHEN Q L, et al. Agricultural technological progress, rural labor force transfer and farmers’ income — Analysis of grouped PVAR model based on agricultural labor productivity[J]. Journal of Agrotechnical Economics, 2017(6): 28−41 [24] 陈雪婷, 黄炜虹, 齐振宏, 等. 生态种养模式认知、采纳强度与收入效应−以长江中下游地区稻虾共作模式为例[J]. 中国农村经济, 2020(10): 71−90CHEN X T, HUANG W H, QI Z H, et al. Farmers ’ cognition, adoption intensity and income effect of ecological breeding mode: a case study of rice-shrimp co-cultivation mode in the middle and lower reaches of Yangtze River[J]. Chinese Rural Economy, 2020(10): 71−90 [25] 侯晓康, 刘天军, 黄腾, 等. 农户绿色农业技术采纳行为及收入效应[J]. 西北农林科技大学学报(社会科学版), 2019, 19(3): 121−131HOU X K, LIU T J, HUANG T, et al. Adoption behavior and income effects of green agricultural technology for farmers[J]. Journal of Northwest A& F University (Social Science Edition), 2019, 19(3): 121−131 [26] 耿宇宁, 郑少锋, 刘婧. 农户绿色防控技术采纳的经济效应与环境效应评价−基于陕西省猕猴桃主产区的调查[J]. 科技管理研究, 2018, 38(2): 245−251 doi: 10.3969/j.issn.1000-7695.2018.02.036GENG Y N, ZHENG S F, LIU J. The evaluation of economic and environmental effects of farmers’ integrated pest management technology adoption behavior: Based on the investigation in kiwi fruit main production areas of Shaanxi Province[J]. Sicence and Technology Management Research, 2018, 38(2): 245−251 doi: 10.3969/j.issn.1000-7695.2018.02.036 [27] 熊鹰, 何鹏. 绿色防控技术采纳行为的影响因素和生产绩效研究−基于四川省水稻种植户调查数据的实证分析[J]. 中国生态农业学报(中英文), 2020, 28(1): 136−146XIONG Y, HE P. Impact factors and production performance of adoption of green control technology: an empirical analysis based on the survey data of rice farmers in Sichuan Province[J]. Chinese Journal of Eco-Agriculture, 2020, 28(1): 136−146 [28] 周洁红. 农户蔬菜质量安全控制行为及其影响因素分析−基于浙江省396户菜农的实证分析[J]. 中国农村经济, 2006(11): 25−34ZHOU J H. Analysis on farmers’ control behavior of vegetable quality and safety and its influencing factors — Based on the empirical analysis of 396 vegetable farmers in Zhejiang Province[J]. Chinese Rural Economy, 2006(11): 25−34 [29] DIMARA E, SKURAS D. Adoption of agricultural innovations as a two-stage partial observability process[J]. Agricultural Economics, 2003, 28(3): 187−196 doi: 10.1111/j.1574-0862.2003.tb00137.x [30] VAN DE VEN W P M M, VAN PRAAG B M S. The demand for deductibles in private health insurance: A probit model with sample selection[J]. Journal of Econometrics, 1981, 17(2): 229−252 doi: 10.1016/0304-4076(81)90028-2 [31] 秦诗乐, 吕新业. 农户绿色防控技术采纳行为及效应评价研究[J]. 中国农业大学学报(社会科学版), 2020, 37(4): 50−60QIN S L, LYU X Y. Research on farmers’ green control techniques adoption behavior and its effect evaluation[J]. Journal of China Agricultural University (Social Sciences), 2020, 37(4): 50−60 [32] LOKSHIN M, SAJAIA Z. Maximum likelihood estimation of endogenous switching regression models[J]. The Stata Journal: Promoting Communications on Statistics and Stata, 2004, 4(3): 282−289 doi: 10.1177/1536867X0400400306