Evaluation of Government-Subsidized Agricultural Insurance for Poverty Alleviation: Evidence from a Quasi-natural Experiment in China
By Duan Baige and He Minhua
In The Public Interest, May 2022
This article was originally published on the WeChat account of the China Association of Actuaries on Feb. 9, 2022
(https://mp.weixin.qq.com/s/PWZm1of6sGnn9RRnKEZkTA). The full version has been published in CSSCI Journal Insurance Research, issue 11, 2021.
Poverty significantly affects human development and well-being, and the fight against it is an important and historical task facing many countries in the world. Theoretically, poverty, as the core issue of development economics research, has been the focus of economists. Amartya Sen, the 1998 Nobel laureate in Economic Sciences, pointed out in his book Poverty and Famine that “the real cause of poverty is the lack of capacity and opportunity of the poor to generate income. In essence, it is a poverty of capabilities and rights.”[1] Empirically, as one of the three major battles concerning China’s high-quality economic development, it is of great importance to scientifically assess the practical effects of poverty alleviation and eradication policy tools. Some studies have focused on the poverty alleviation and reduction effects of policy tools such as the establishment of the United Nations’ International Fund for Agricultural Development (IFAD), state poverty counties and pilot poverty alleviation reform zones, but systematic empirical studies still need to be conducted.
Background
Risk is an important constraint on economic development, affecting both economic development itself (“making the cake bigger”) and the quality of economic development (including common prosperity). Data from the State Council’s Poverty Alleviation Office in 2017 show that disasters are the second leading cause of poverty in China (20 percent), with diseases being the leading cause. Agricultural insurance can effectively spread the risk of agricultural production and operation and thus help farmers resist the impact of major disasters. It has become a common agricultural risk management tool around the globe. Agricultural insurance systems have been established and implemented in more than 100 countries worldwide. China started to implement the government-subsidized agricultural insurance system in 2007; it initially launched the pilot work of subsidizing agricultural insurance premiums from the central government in six pilot provinces—Inner Mongolia, Jilin, Jiangsu, Hunan, Xinjiang and Sichuan. By the end of 2012, 31 provinces (autonomous regions and municipalities directly under the central government) had launched the pilot work, which promoted the rapid development of agricultural insurance and significantly increased the coverage rate. China’s government-subsidized agricultural insurance is an efficient policy of the central and local treasuries to provide insurance premium subsidies for agricultural insurance that meets the requirements and is based on the principle of autonomy and voluntary participation. Due to the vast area of China, there are huge differences in the level and characteristics of agricultural development, the varieties of advantageous agricultural products and the main agricultural risks faced by different regions. Currently, there is no specific unified standard for agricultural insurance in China, and local authorities at all levels have the right to formulate specific implementation plans in conjunction with local conditions, which are monitored and inspected by the central and higher authorities.
As far as the sharing of premiums is concerned, the financial department (central and local) is responsible for most of the premiums, with farmers’ own premiums generally not exceeding 50 percent, depending on the region and type of insurance.
In terms of what can be insured, China’s government-subsidized agricultural insurance is basically divided into centrally subsidized insurance, provincial subsidized insurance, municipal subsidized insurance, district subsidized insurance and so on, in accordance with the principle of high-level financial support for bulk products and low-level financial support for special products. Centrally subsidized insurance mainly covers the major bulk agricultural products that are related to the national economy, people’s livelihood, food and ecological security of the country. Provincial subsidized insurance is mainly for local, large-volume agricultural products in line with local agricultural production.
Insurance generally covers major natural disasters, major pests and diseases, and accidents with differentiated implementation plans according to local needs. Cost insurance, income insurance and index insurance—whose liabilities are changes in prices, yields and weather—are also being actively piloted in various regions.
Methodology and Results
Has the implementation of government-subsidized agricultural insurance, an inclusive financial tool, had any effect on poverty alleviation? For instance, has it alleviated rural poverty? If so, to what extent?
To investigate this, our research used the ideal “quasi-natural experiment” of government-subsidized agricultural insurance premium subsidies, piloted in different regions of China since 2007, to test the policy effects and transmission mechanisms of implementing such insurance on regional poverty alleviation. We employed a multi-period difference-in-differences (DID) approach to answer the question empirically and scrupulously. Specifically, we identified the causal effects between the government-subsidized agricultural insurance pilot and the incidence, depth and intensity of poverty among rural residents in each region.
This study found that the effect of government-subsidized agricultural insurance on poverty alleviation and reduction is significant, with an average decrease of 0.016 in poverty incidence, 0.008 in poverty depth and 0.005 in poverty intensity after implementation of the insurance. This indicates that government-subsidized agricultural insurance not only reduces the number of poor people by about 160 per 10,000, but also narrows the relative poverty gap and internal income disparity of poverty-level farm households.
Several types of tests were used to guarantee the validity and thoroughness of the results. The parallel trend test with an extended sample period ensured the randomness of the experimental and control groups in terms of grouping and events. The placebo test with a constructed counterfactual analysis further validated and supported the robustness of the underlying regression model results. A heterogeneity analysis of farmers’ income structure and average household size was also performed. This helped us to better understand and identify that increasing farmers’ household business income is an important transmission channel and influence mechanism for allowing agricultural insurance to exert its poverty alleviation and reduction effects. The heterogeneity analysis of farmers’ income structure and the test of transmission mechanism showed that after the implementation of government-subsidized agricultural insurance, the poverty incidence, depth and intensity of farm households with a high proportion of business income decreased by an average of 0.084, 0.045 and 0.027, respectively. Its contribution to the decrease of the three poverty indicators reached more than 50 percent.
Agricultural insurance has an even greater role in alleviating farmers’ poverty. The per capita farmers’ household business income increased by at least 4.5 percent on average. The heterogeneity analysis of average household area and the transmission mechanism test showed that the poverty alleviation and reduction effect of agricultural insurance was more obvious for farmers with larger average household areas. Average household acreage had a mediating effect on poverty alleviation and reduction by increasing household business income.
Discussion
Our research shows, firstly, that the implementation of policy-based agricultural insurance has been significantly effective in reducing poverty incidence, depth and intensity. Over the years, as the central and local governments have continued to increase agricultural insurance premium subsidies, optimize the premium subsidy structure, expand the scope of agricultural products, and promote more cities and counties in each province to carry out pilot projects, the effect of poverty alleviation and reduction has become more and more obvious.
Secondly, in the new round of agricultural insurance pilots, more detailed incentive mechanisms and programs should be designed based on household-level heterogeneity, such as farmers’ income structure and average household (arable) area, taking into account the risks for different crops. Targeted and differentiated premium subsidy policies should be formulated to avoid weakening the poverty alleviation effect due to the burden of premium expenditure. Attention should also be paid to the determinants affecting poverty, distinguishing between risk-based poor and capital-based poor farmers. Improved precision of poverty alleviation and reduction efforts will help extremely poor farmers escape the poverty caused by disasters, consolidating and expanding poverty eradication.
Thirdly, as the main body responsible for government-subsidized agricultural insurance, insurance companies should not only continue to promote the expansion of agricultural insurance, increase the number of products and raise the standard of coverage, but also continue to explore their comparative advantages in professional technology and supervision mechanisms. In addition to providing farmers with professional technical support such as risk management, companies need to assess the different risk types of insured farmers, motivate the insured to reduce risk, reduce adverse selection and moral hazards, and promote the high-quality development of agricultural insurance.
The innovations and contributions of this research are twofold. First, in terms of empirical identification strategy, this is the first time that the causal identification method of multi-period DID has been used to test and evaluate the effect of China’s government-subsidized agricultural insurance on precise poverty alleviation along with its transmission channel and influence mechanism. Second, in terms of poverty index measures, we selected three common poverty indices—rural residents’ poverty incidence, depth and intensity—which makes the causal identification more detailed and comprehensive.
This research also has important policy implications. First, it uses policy-based agricultural insurance as an example to provide a set of research paradigms to test the practical effects of implementing policy instruments that combine effective markets and responsive government. Second, as common prosperity is an appropriate part of high-quality economic development, our research also provides an academic basis and insight for understanding how insurance can be used to address income risk, alleviate rural poverty and solve the problems facing common prosperity. Agricultural insurance, in addition to government redistribution, may prove to be a significantly effective market solution to those problems.
Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries, the editors, or the respective authors’ employers.
Duan Baige, Ph.D., is an associate professor and doctor of the School of Economics, at Fudan University. Duan can be reached at duanbaige@fudan.edu.cn.
He Minhua is a student majoring in finance from a joint program of the School of Economics, Fudan University and the London School of Economics and Political Science. She can be reached at manwah.ho@outlook.com.