Research

INTRODUCTION

Pastoral livestock production is practiced in 25% of the global land area (Nori et al. 2008). Rangelands provide 10% of the global meat supply and support an estimated 200 million pastoralists’ households and the herds of nearly 1 billion camelids, cattle, and smaller livestock, in addition to yaks, horses, reindeer, and other ungulates (Neely et al. 2009). Generally, pastoral livestock occurred in places where constraining soil, rainfall, and temperature conditions render the land unsuitable for crop cultivation, like arid regions (sub-Sahara and central China) and high-cold regions (Qinghai-Tibet Plateau, QTP) with the characteristic of limited, highly variable, and unpredictable resource endowment. Furthermore, the death of animals leads to permanent asset erosion and can have longer-term consequences than the seasonal loss of income resulting from a failed crop (Ruchismita and Churchill 2012). As a result, pastoral livestock is more sensitive to the natural environment compared to other agricultural sections.

Insurance is an important formal mechanism to manage the risk in the pastoral areas under adverse weather conditions and similar events beyond the control of primary producers. There is an increasing interest among policy makers and practitioners to use insurance as a financial mechanism to mitigate the negative impacts of risk in agriculture. Within China, grasslands constitute roughly 40% of the total land area, spanning approximately 400 million hectares. The Chinese government launched a livestock insurance program in the pastoral areas in 2011 with the aim of assisting herdsmen in alleviating risks associated with their livestock production (Liu et al. 2021). However, the effects of this program on livestock production and its potential ecological impact are still not well comprehended. A sufficient understanding of this subject is of utmost importance for the successful implementation and expansion of livestock insurance policies.

Yet it is surprising that the result of previous studies of the effects of agricultural insurance on pastoralist’s welfare is inconsistent. Some studies revealed that agricultural insurance yielded substantial economic and social benefits, such as stabilizing income risk (Akter et al. 2017), promoting productivity (Zou et al. 2022) , increasing investment in production technologies (Fang et al. 2021), poverty alleviation (Chantarat et al. 2013), and encouraging sustainable agriculture practices (Möhring et al. 2020). However, other studies indicate that farmers’ enthusiasm for input factors may decrease because of moral hazards and adverse selection effects (Norton et al. 2016, Rao and Zhang 2020), and increased use of chemical fertilizers and pesticides have been observed (Bosch et al. 2017, Weber et al. 2016), which can lead to adverse environmental consequences.

To address this research gap, this study conducted a comprehensive analysis of the impact of livestock insurance on production in the QTP, specifically focusing on livestock income and input factors. We utilized a survey dataset that was collected from 216 households in the QTP region to analyze the impact of livestock insurance on production. To mitigate the potential influence of self-selection bias, we employed the Propensity Score Matching (PSM) approach for correction.

STUDY AREA AND DATA

Study area

Qinghai-Tibet Plateau (QTP) is China’s largest pastoral area, located in the southwest part of the country (Fig. 1). The grassland area of the QTP is the largest among the seven pastoral regions in China, an area of about 165.38×10⁴ km², accounting for 41.88% of the total grassland area in China (Zhang et al. 2016). More than 80% of the Tibetan Plateau is above 3000-meter elevation, sometimes known as the Third Pole.

Grazing is a high-risk enterprise on the Tibetan plateau, where high elevation, long winters, and severe snowstorms are normal features of the grazing environment. Traditionally, herd and family mobility are the best and the most important way of using scarce resources heterogeneously distributed in space and time (Nozières et al. 2011). However, the privatization of pastoral lands and the associated trend toward sedentarization of pastoralists in developing countries reduced the mobility of animals. Livestock insurance is an important formal mechanism to manage the risk in pastoral areas under adverse weather conditions and similar events beyond the control of primary producers.

The policy-oriented livestock insurance program was introduced in the QTP in 2011. The government offers a 90% subsidy on premiums for livestock insurance in the QTP (Li et al. 2021). Our field survey data indicate that the premium for yaks and sheep is 15 and 2.5 CN¥ per head per year, respectively. Consequently, the government subsidizes the premiums of 135 and 22.5 CN¥ per head per year for yaks and sheep, respectively. In the event of the insured animal’s death within a year, households receive indemnities of 3000 CN¥ for a yak and 500 CN¥ for a sheep. After a decade of implementing livestock insurance, it is crucial to comprehensively assess the impact of this insurance product.

Sample characteristics

This study used herder household survey data from the QTP in 2019. Tibet’s grasslands are categorized into four distinct zones based on climate and livestock suitability: warm and humid forest scrub (southeast); cold and semi-humid meadow (northeast); cold and arid-semiarid (south-central), and arid-desert (northwest; The Team of Comprehensive Scientific Expedition on the Qinghai-Tibet Plateau 1992). Because of the sparse population in the arid-desert zone, we focused our field study on the other three zones.

We employed a stratified random sampling method. First, we randomly selected two counties from each grassland zone, totaling six counties: Gongbujiangda County and Yadong County (warm and humid forest scrub zone), Baqing County and Dingqing County (cold and semi-humid meadow zone), and Zhongba County and Bange County (cold and arid-semiarid zone). Next, within each county, we randomly selected two towns based on the livestock production scale, choosing one town with larger and one with smaller household grassland areas. From each town’s list of villages, we then randomly selected two villages. Finally, with the assistance of local government, we randomly selected six herder households from each village to ensure a representative sample across varying livestock production levels. In total, 216 rural households across 36 villages in 12 townships in 6 counties were randomly selected and face-to-face interviewed in August 2019. The survey collected data on pastoral households’ production, income, demography, and grassland use in 2018. The spatial distribution of sampling counties is shown in Figure 1.

A summary of surveyed variables is presented in Table 1. It shows that 83.3% (180/216) of herder households participated in livestock insurance. The high participation rate can be attributed to the significant subsidies for policy-oriented livestock insurance. Most herders who purchased livestock insurance were repeat users. Nearly all (167/169, 98.8%) herders who purchased insurance in 2017 also purchased insurance in 2018. Notably, only 7.2% (11/180) of herders were new insurance purchasers. The surveyed household heads had an average age of approximately 53 years and a mean educational attainment of 1.685 years.

The average contracted grassland area per household is 5286 mu, which is equivalent to 352.4 ha. The average livestock income was 15110 CN¥ in 2018. The total number of person-months dedicated to livestock work is 17.546. The average annual input for livestock production is 6170 CN¥, and the average percentage of livestock offtake is 6.185%.

Table 2 presents the mean differences in livestock income and production characteristics with and without livestock insurance. Although statistically similar across most observed characteristics, the two groups differed in livestock sales. The group with livestock insurance exhibited significantly higher sales (7.05% vs 1.85%). However, these differences did not have the effects of other factors, especially the factors that affect herders’ decision to be involved in livestock insurance, so the information presented in Table 2 is inconclusive. More rigorous and robust assessments are needed.

METHODS

To estimate the impact of livestock insurance on herder households, the model constructed in this paper is as follows:

(1)

Where Y_i represents the livestock production of household i (livestock income, productive inputs, etc.). D_i represents whether the household has livestock insurance. If the household has livestock insurance, D_i = 1, otherwise, D_i = 0. X_i are the control variables, and α₁, α₂, and α₃ are the coefficients to be estimated. ε_i is the random disturbance term.

Because households were not randomly assigned to the two groups (with and without livestock insurance), whether households purchase livestock insurance is not random behavior. It is impossible to observe the characteristics obtained by the households who purchased the livestock insurance if they did not purchase it. Therefore, the ordinary least squares (OLS) method may lead to issues of bias. To solve the self-selection problem, the widely used propensity score matching (PSM) method is adopted in this study.

The basic idea of PSM is to find a control group with similar characteristics to the household who participated in the livestock insurance among the herders who did not participate. PSM uses the logit model to estimate the probability of their participation and constructs an approximately randomized counterfactual data to compare the impact of livestock insurance on households’ income and production behaviors.

First, we constructed a decision model for households to participate in livestock insurance and used the Logit model to estimate the possibility of households’ participation in livestock insurance (propensity scores of participating).

(2)

The dummy variable D_i is used to indicate whether the individual household i participates in the livestock insurance, and “1” means to participate (treatment group), and “0” means not to participate (control group). X_i are the control variables selected after matching.

After obtaining the propensity matching score, an appropriate matching method was selected to match the participants and non-participants. Commonly used matching methods include kernel matching, K-nearest neighbor matching, caliper matching, and local linear regression matching (Tang et al. 2022, Zhang et al. 2024). In this study, different matching methods were used to verify the robustness of the estimated results. When the matching was completed, the effect of livestock insurance on production could be measured by the average treatment on the treated effect (ATT).

(3)

Where Y_i¹ can be directly observed and represents the livestock production of herders who have livestock insurance. Y_i⁰ is not observable. It is a counterfactual result constructed based on propensity score matching among households who did not participate in livestock insurance, indicating the household income and production input of herders who have livestock insurance if they did not have livestock insurance.

RESULTS AND DISCUSSION

The results of matching

The first step of PSM involved estimating the selection model. We controlled for factors influencing livestock insurance uptake based on prior research. Premium rates and subsidies, although crucial elsewhere (Cole et al. 2013), were excluded because they were uniform for herders on the QTP. We included factors like grassland area (wealth and production indicator) (Hill et al. 2013), smartphone usage (ICT adoption; Ali et al. 2020), experience with disasters (Arshad et al. 2016, Michel-Kerjan and Kousky 2010, Liu et al. 2019), the cost of purchased forage for supplementary feeding and hayfield (Goodrich et al. 2020, Yang et al. 2021), and participation in social networks through cooperatives and other insurance (Cole et al. 2014, Hill et al. 2016, Karlan et al. 2014). Specifically, we controlled for household labor force because it is a crucial factor influencing both the scale of livestock production and the household’s capacity to implement grassland management practices (Tan et al. 2018). Households with a larger labor force may have a greater capacity for livestock rearing and may therefore be more likely to adopt insurance as a risk management tool. Additionally, we controlled for household characteristics like age, education, and household size.

The result of the decision model using the logit model is presented in Table 3. The results show that six factors have a statistically significant impact on the demand for livestock insurance: household size, cooperative membership, prior disaster experience, agricultural subsidies, health insurance purchasing history, and county of residence.

Four matching methods were used in this research, i.e., kernel matching, k-nearest neighbor matching, k-nearest-neighbor matching within caliper, and caliper matching. As Figure 2 reveals, the participants and non-participants have a larger common support area (that is, the overlap interval of the propensity scores of the treatment group and the control group) after matching. This means our matching satisfied the common support area test. Table 4 presents the results of the matching balance test with different methods. The pseudo R², LR chi², mean and median biases declined significantly after matching (a detail results of balancing test are represented in Appendix 1). Therefore the matching balance test was passed.

The treatment effects of livestock insurance

The estimated results of the treatment effects of livestock insurance on production are shown in Table 5. OLS and four matching methods were used. The detailed OLS estimation results are presented in Appendix 2. To assess the sensitivity of our findings to alternative model specifications, we calculated Oster bounds (Oster 2019, Sangwan and Kumar 2021), which are reported in Appendix 3. These results indicate that our estimates are unlikely to be significantly affected by omitted variable bias.

As revealed in Table 5, there are significant differences in livestock production between insured and uninsured households. Consistent results across various methods support the robustness of our findings.

Table 5 demonstrates a strong correlation between livestock insurance participation and increased livestock income. This association persists across different matching methods used for estimation. Taking the kernel matching method as an example, insured households reported an average livestock income of 16,330 CN¥, significantly higher than the 5690 CN¥ of uninsured households. Across all four matching methods, the mean ATT is estimated at 10,900 CN¥, indicating a 198.9% increase in livestock income from 5480 CN¥ to 16,380 CN¥ for insured households compared to those without livestock insurance. This finding aligns with previous research conducted by Bhuiyan et al. (2022) and Tan et al. (2022), where they similarly observed that insurance policies contributed to an increased agricultural income. Such a result is expected, given the significant government subsidies of up to 90% that policy-oriented livestock insurance in the QTP receives. Consequently, households can undoubtedly reap significant benefits from participating in this program. This finding highlights the crucial role of existing policy-oriented livestock insurance in augmenting household income in the QTP region.

Higher livestock income is observed among households participating in livestock insurance programs (Table 5). This correlation might be linked to increased investment in production activities. As the table suggests, livestock insurance is associated with higher levels of labor input in livestock production. The mean ATT of livestock insurance on labor input in livestock production is estimated to be 5.848 person-months. This means the household’s livestock labor would increase by 5.848 person-months after participating in livestock insurance. One potential drawback with agricultural insurance is its association with moral hazard, which has the potential to hinder production (Regmi et al. 2022). Moral hazard refers to the possibility that insured individuals may take fewer precautions against potential harm compared to those without insurance (Horowitz and Lichtenberg 1993). This can manifest in various ways, including reduced resource and input utilization. In the context of the QTP, known for its unique ecological characteristics and agricultural practices, understanding the implications of moral hazard is crucial. However, our research findings indicate that livestock insurance in the QTP does not give rise to moral hazard issues, instead, it promotes investment in livestock production. This finding is consistent with the previous literature (Chang and Mishra 2012, Lenel and Steiner 2020, Sibiko and Qaim 2020).

It is important to note that capital input did not show a significant difference between the two groups. This finding highlights the differential impacts of livestock insurance on labor and capital investments. The disparity can be attributed to the existing capital allocation of households, which is constrained by their limited financial resources. Our survey data reveals that the top third of households by income invested significantly more capital, averaging 7980.8 CN¥, compared to 5636.4 CN¥ invested by the bottom third. High-income households have already committed more capital, whereas low-income households, constrained by limited income, are unable to make additional investments. Consequently, the purchase of insurance did not have a consistent impact on capital investment across the overall sample of households. This suggests that the benefits of insurance vary across income groups, a topic we will discuss further in the heterogeneous analysis.

As mentioned earlier, the QTP experiences low livestock sale rates. This results not only in low income but also heavy grazing pressure. Overgrazing has been widely recognized as a significant factor contributing to grassland degradation in the QTP (Tao and Gao 1992, Harris 2010, Zhao et al. 2018, Yang et al. 2023). The ecological benefits of livestock insurance are noteworthy, particularly in the context of reducing grazing pressure and mitigating grassland degradation. To further investigate the potential ecological effects of livestock insurance, we also analyzed its influence on livestock sales (see Table 5). The statistical significance was determined using PSM, providing robust evidence for the relationship. The results demonstrate that livestock insurance has a significantly positive effect on household livestock sales. The mean of ATT is estimated to be 6.483, indicating that participating in livestock insurance leads to an increase of 6.483 percentage points in livestock sales (from 0.909% to 7.392%). According to our field survey data, households in the top 50% of livestock offtake exhibit a significantly higher rate of livestock insurance adoption (88.9%) compared to other households (77.8%). Furthermore, our analysis shows that repeated buyers have substantially higher livestock sale rates (7.41%) than new buyers (2.51%), thereby providing further support for the hypothesis that insurance incentivizes increased livestock offtake. Increasing the sale of livestock results in higher income from livestock. Furthermore, the increased off-take of livestock led to a decrease in the livestock population grazing on grasslands. Consequently, the implementation of livestock insurance in this context can generate unforeseen ecological advantages, highlighting the significance of adopting livestock insurance as a strategy to address this environmental issue.

This finding contradicts the results of Gebrekidan et al. (2019), which indicated that households that purchased livestock insurance are less likely to sell their herds in Ethiopia. It is important to note that the QTP and Ethiopian contexts exhibit distinct regional characteristics, which may contribute to the varied results. These differences could include variations in risk response among pastoralists, as influenced by cultural, economic, or environmental factors.

In the QTP, herders traditionally consider the size of their herds as a measure of their wealth, leading them to opt for retaining a larger number of livestock on the grassland while slaughtering or selling fewer animals (Miller 2005, Bai et al. 2008). This belief stems from the notion that the accumulating livestock can act as a precautionary saving strategy to hedge against risk (Nyima 2014, Yeh et al. 2017). In situations where formal insurance options are limited or unavailable, relying on livestock accumulation as a form of self-insurance becomes a practical approach to managing risk in pastoral communities (Næss and Bårdsen 2010, 2013, Johannesen and Skonhoft 2011). By accumulating livestock, herders create a valuable asset that can serve as a buffer against the potential risk of substantial herd loss, thereby safeguarding their livelihoods and ensuring resilience against challenging winter conditions (Shang et al. 2012, 2014). This practice aligns with the principles of the precautionary saving theory, which suggests that individuals or households invest in assets that can be used to cope with future uncertainties. Herd accumulation is an ex-ante risk mitigation strategy in QTP, and the livestock insurance transfers productive risks outside the farm, reducing the need to maintain an excessive herd size. Therefore, participating herders are more willing to sell livestock. In contrast, in the Ethiopian context, herd off-take is a major ex-post risk coping strategy commonly employed by pastoralists (Gebrekidan et al. 2019, Ngigi et al. 2015). These herders commonly engage in selling livestock to buffer consumption during uncertainty and economic stress, such as prolonged drought. When households have the assurance provided by a formal insurance contract, they tend to curtail or discontinue the practice of adverse herd off-take as a mechanism for coping with risks.

Heterogeneous analysis

To examine the heterogeneous effects of insurance on livestock production at different levels of income, we conducted a heterogeneity test with the K-nearest neighbor matching method. We divided the sample equally into three groups according to income level. The results are presented in Table 6. The results show that livestock insurance has significant impacts on low- and medium-income groups but little impact on high-income groups. It is important to note that while purchasing insurance does not significantly impact capital investment for the overall sample of herder households, it does have a significant effect on capital investment among low-income herder households.

The possible explanation is that the wealthier households already have more ways to deal with production risk, e.g., formal credit. In the pastoral China, the herders could obtain loans from formal institutions when facing financial difficulties (Zhang et al. 2022). However, formal credit is more accessible to the wealthier households (Hermes and Lensink 2011, Zhang et al. 2024). This suggests that insurance may act as a mechanism that alleviates some of the financial constraints faced by these households, enabling them to make additional investments in livestock production that they might not have been able to afford otherwise. Although low-income households may have limited capital initially, the security provided by insurance can potentially encourage and enable them to invest more in their livestock. This implies that livestock insurance in the QTP has the potential to benefit the poor by facilitating their productive investments and contributing to reducing the income disparity among herders in QTP.

CONCLUSION AND IMPLICATIONS

This study empirically investigated the effects of livestock insurance on livestock production in the Qinghai-Tibetan Plateau (QTP) based on a dataset of 216 households. To address self-selection bias, we employed the Propensity Score Matching (PSM) approach. Furthermore, we investigated the heterogeneous effects of livestock insurance on different household income level.

The results of this study show that livestock insurance exerts a positive and statistically significant impact on livestock production in the QTP. It is observed that livestock insurance does not give rise to moral hazard concerns; on the contrary, it stimulates investment in livestock production, leading to an increase in household livestock income. On average, the insured households allocated 5.848 person-months more toward livestock production compared to uninsured households, and this was associated with a livestock income increase of 10,900 CN¥. Overall, the evidence presented in this paper highlights the favorable economic effects of livestock insurance in the QTP.

Furthermore, our research results indicate potential ecological benefits associated with livestock insurance. Traditionally, pastoralists in the QTP have exhibited reluctance to sell their animals, which has often translated into the maintenance of large herd sizes and, consequently, the potential for overgrazing of grasslands. By reducing production risks faced by households, livestock insurance encourages an increase in off-take rates. This means the livestock insurance has the potential to alleviate grazing pressure and contribute to the conservation of grasslands in the QTP.

Our findings also reveal equity impacts of livestock insurance. Heterogeneous analysis shows that the low-income households derive greater benefits from livestock insurance, while the livestock production of high-income households is not significantly affected. This suggests that livestock insurance can contribute to poverty alleviation and the reduction of income disparity in the QTP.

The policy implications of our results are significant. Given the dual ecological and well-being benefits of livestock insurance, policy makers should continue to promote policy incentives to encourage participation in the livestock insurance program among herders in the QTP. Specifically, providing additional premium subsidies for low- and medium-income households to join livestock insurance programs is deemed necessary.

This study sheds light on the relationship between livestock insurance and production outcomes, but limitations are worth noting. PSM effectively addresses selection bias from unobserved variables. However, the estimated ATT effects may vary depending on the specific characteristics of the sampled pastoral households. Future research could employ alternative methods to estimate the causal effects of participation in livestock insurance. Additionally, the cross-sectional nature of our data limits our ability to track changes in livestock production over time. This dynamic aspect warrants further investigation in future studies.

LITERATURE CITED

Akter, S., T. J. Krupnik, and F. Khanam. 2017. Climate change skepticism and index versus standard crop insurance demand in coastal Bangladesh. Regional Environmental Change 17:2455-2466. https://doi.org/10.1007/s10113-017-1174-9

Ali, W., A. Abdulai, and A. K. Mishra. 2020. Recent advances in the analyses of demand for agricultural insurance in developing and emerging countries. Annual Review of Resource Economics 12(1):411-430. https://doi.org/10.1146/annurev-resource-110119-025306

Arshad, M., T. S. Amjath-Babu, H. Kächele, and K. Müller. 2016. What drives the willingness to pay for crop insurance against extreme weather events (flood and drought) in Pakistan? A hypothetical market approach. Climate and Development 8(3):234-244. https://doi.org/10.1080/17565529.2015.1034232

Bai, J. H., H. Ouyang, B. S. Cu, Q. G. Wang, and H. Chen. 2008. Changes in landscape pattern of alpine wetlands on the Zoige plateau in the past four decades. Acta Ecologica Sinica 28(5):2245-2252. https://doi.org/10.1016/S1872-2032(08)60046-3

Bhuiyan, M. A., M. Davit, Z. XinBin, and Z. Zurong. 2022. The impact of agricultural insurance on farmers’ income: Guangdong Province (China) as an example. PLoS ONE 17(10):e0274047. https://doi.org/10.1371/journal.pone.0274047

Bosch, J., A. Rodríguez, I. Iglesias, M. J. Muñoz, C. Jurado, J. M. Sánchez-Vizcaíno, and A. Torre. 2017. Update on the risk of introduction of African swine fever by wild boar into disease-free European Union countries. Transboundary Emerging Diseases 64:1424-1432. https://doi.org/10.1111/tbed.12527

Chang, H. H., and A. K. Mishra. 2012. Chemical usage in production agriculture: do crop insurance and off-farm work play a part? Journal of Environmental Management 105:76-82. https://doi.org/10.1016/j.jenvman.2012.03.038

Chantarat, S., A. G. Mude, C. B. Barrett, and M. R. Carter. 2013. Designing index-based livestock. insurance for managing asset risk in northern Kenya. Journal of Risk Insurance 80(1):205-237. https://doi.org/10.1111/j.1539-6975.2012.01463.x

Cole, S., X. Gine, J. Tobacman, P. Topalova, R. Townsend, and J. Vickery. 2013. Barriers to household risk management: evidence from India. American Economic Journal: Applied Economics 5(1):104-135. https://doi.org/10.1257/app.5.1.104

Cole, S., D. Stein, and J. Tobacman. 2014. Dynamics of demand for index insurance: evidence from a long-run field experiment. American Economic Review 104(5):284-290. https://doi.org/10.1257/aer.104.5.284

Fang, L., R. Hu, H. Mao, and S. J. Chen. 2021. How crop insurance influences agricultural green total factor productivity: evidence from Chinese farmers. Journal of Cleaner Production 321:128977. https://doi.org/10.1016/j.jclepro.2021.128977

Gebrekidan, T., Y. Guo, S. Bi, J. Wang, C. Zhang, J. Wang, and K. Lyu. 2019. Effect of index-based livestock insurance on herd offtake: evidence from the Borena zone of southern Ethiopia. Climate Risk Management 23:67-77. https://doi.org/10.1016/j.crm.2018.10.003

Goodrich, B., J. Yu, and M. Vandeveer. 2020. Participation patterns of the rainfall index insurance for pasture, rangeland and forage programme. Geneva Papers on Risk and Insurance - Issues and Practice 45:29-51. https://doi.org/10.1057/s41288-019-00149-3

Harris, R. B. 2010. Rangeland degradation on the Qinghai-Tibetan plateau: a review of the evidence of its magnitude and causes. Journal of Arid Environment 74:1-12. https://doi.org/10.1016/j.jaridenv.2009.06.014

Hermes, N., and R. Lensink. 2011. Microfinance: its impact, outreach, and sustainability. World Development 39:875-881. https://doi.org/10.1016/j.worlddev.2009.10.021

Hill, R. V., J. Hoddinott, and N. Kumar. 2013. Adoption of weather - index insurance: learning from willingness to pay among a panel of households in rural Ethiopia. Agricultural Economics 44(4-5):385-398. https://doi.org/10.1111/agec.12023

Hill, R. V., M. Robles, and F. Ceballos. 2016. Demand for a simple weather insurance product in India: theory and evidence. American Journal of Agricultural Economics 98(4):1250-1270. https://doi.org/10.1093/ajae/aaw031

Horowitz, J. K., and E. Lichtenberg. 1993. Insurance, moral hazard, and chemical use in agriculture. American Journal of Agricultural Economics 75(4):926-935. https://doi.org/10.2307/1243980

Johannesen, A. B, and A. Skonhoft. 2011. Livestock as insurance and social status: evidence from reindeer herding in Norway. Environmental and Resource Economics 48:679-694. https://doi.org/10.1007/s10640-010-9421-2

Karlan, D., R. Osei, I. Osei-Akoto, and C. Udry. 2014. Agricultural decisions after relaxing credit and risk constraints. Quarterly Journal of Economics 129(2):597-652. https://doi.org/10.1093/qje/qju002

Lenel, F., and S. Steiner. 2020. Formal insurance and solidarity: experimental evidence from Cambodia. Journal of Economic Behavior & Organization 174:212-234. https://doi.org/10.1016/j.jebo.2020.02.002

Li, F. N., C. S. Qin, and X. Zhou. 2021. Design of a snow disaster weather index insurance for alpine pastoral areas: a case study in Gannan. Acta Prataculturae Sinica 30:199-204. [Translated from the Chinese.]

Liu, P. F., L. L. Hou, D. Q. Li, S. Min, and Y. Y. Mu. 2021. Determinants of livestock insurance demand: experimental evidence from Chinese herders. Journal of Agricultural Economics 72(2):430-451. https://doi.org/10.1111/1477-9552.12402

Liu, X., Y. Tang, J. Ge, and M. J. Miranda. 2019. Does experience with natural disasters affect willingness-to-pay for weather index insurance? Evidence from China. International Journal of Disaster Risk Reduction 33:33-43. https://doi.org/10.1016/j.ijdrr.2018.09.006

Michel-Kerjan, E. O., and C. Kousky. 2010. Come rain or shine: evidence on flood insurance purchases in Florida. Journal of Risk and Insurance 77(2):369-397. https://doi.org/10.1111/j.1539-6975.2009.01349.x

Miller, D. J. 2005. The Tibetan plateau. Chapter 8 in J. M. Suttie, S. G. Reynolds, and C. Batello, editors. Grasslands of the world. Food and Agriculture Organization of the United Nations, Rome, Italy.

Möhring, N., T. Dalhaus, G. Enjolras, and R. Finger. 2020. Crop insurance and pesticide use in European agriculture. Agricultural Systems 184(12):102902. https://doi.org/10.1016/j.agsy.2020.102902

Næss, M. W., and B. J. Bårdsen. 2010. Environmental stochasticity and long-term livestock viability: herd-accumulation as a risk reducing strategy. Human Ecology 38:3-17. https://doi.org/10.1007/s10745-009-9288-4

Næss, M. W., and B. J. Bårdsen. 2013. Why herd size matters—mitigating the effects of livestock crashes. PLoS ONE 8(8):e70161. https://doi.org/10.1371/journal.pone.0070161

Neely, C., S. Bunning, and A. Wilkes. 2009. Review of evidence on drylands pastoral systems and climate change - implications and opportunities for mitigation and adaptation. Food and Agriculture Organization of the United Nations, Rome, Italy.

Ngigi, M., U. Mueller, and R. Birner. 2015. The role of livestock portfolios and group-based approaches for building resilience in the face of accelerating climate change: an asset-based panel data analysis from rural Kenya. ZEF Discussion Papers on Development Policy, No. 205. University of Bonn, Center for Development Research (ZEF), Bonn, Germany. https://doi.org/10.22004/ag.econ.210703

Nori, M., M. Taylor, and A. Sensi. 2008. Browsing on fences: pastoral land rights, livelihoods and adaptation to climate change. International Institute for Environment and Development, Nottingham, UK.

Norton, M., G. J. Sprundel, C. G. Turvey, and M. P. M. Meuwissen. 2016. Applying weather index insurance to agricultural pest and disease risks. International Journal of Pest Management 62(3):195-204. https://doi.org/10.1080/09670874.2016.1184773

Nozières, M. O., C. H. Moulin, and B. Dedieu. 2011. The herd, a source of flexibility for livestock farming systems faced with uncertainties? Animal 5(9):1442-1457. https://doi.org/10.1017/S1751731111000486

Nyima, Y. 2014. A large herd size as a symbol of wealth? Fallacy of the cattle complex theory in Tibetan pastoralism. Area 46(2):186-193. https://doi.org/10.1111/area.12099

Oster, E. 2019. Unobservable selection and coefficient stability: theory and evidence. Journal of Business & Economic Statistics 37(2):187-204. https://doi.org/10.1080/07350015.2016.1227711

Rao. X., and Y. Zhang. 2020. Livestock insurance, moral hazard, and farmers' decisions: a field experiment among hog farms in China. Geneva Papers on Risk and Insurance - Issues and Practice 45(1):134-156. https://doi.org/10.1057/s41288-019-00151-9

Regmi, M., B. C. Briggeman, and A. M. Featherstone. 2022. Effects of crop insurance on farm input use: evidence from Kansas farm data. Agricultural and Resource Economics Review 51(2):361-379. https://doi.org/10.1017/age.2022.5

Ruchismita, R., and C. Churchill. 2012. State and market synergies: insights from India’s microinsurance success. Pages 427-463 in C. Churchill, and M. Matul, editors. Protecting the poor: a microinsurance compendium. Vol. II. International Labour Organization, Geneva, Switzerland.

Sangwan, N., and S. Kumar. 2021. Labour force participation of rural women and the household’s nutrition: panel data evidence from SAT India. Food Policy 102:102117. https://doi.org/10.1016/j.foodpol.2021.102117

Shang, Z. H., M. J. Gibb, F. Leiber, M. Ismail, L. M. Ding, X. S. Guo, and R. J. Long. 2014. The sustainable development of grassland-livestock systems on the Tibetan plateau: problems, strategies and prospects. Rangeland Journal 36(3):267-296. https://doi.org/10.1071/RJ14008

Shang, Z. H., M. J. Gibb, and R. J. Long. 2012. Effect of snow disasters on livestock farming in some rangeland regions of China and mitigation strategies - a review. Rangeland Journal 34:89-101. https://doi.org/10.1071/RJ11052

Sibiko, K. W., and M .Qaim. 2020. Weather index insurance, agricultural input use, and crop productivity in Kenya. Food Security 12:151-167. https://doi.org/10.1007/s12571-019-00987-y

Tan, C., J. Tao, L. Yi, J. He, and Q. Huang. 2022. Dynamic relationship between agricultural technology progress, agricultural insurance and farmers’ income. Agriculture 12:1331. https://doi.org/10.3390/agriculture12091331

Tan. S., R. Zhang, and Z. Tan. 2018. Grassland rental markets and herder technical efficiency: ability effect or resource equilibration effect? Land Use Policy 77:135-142. https://doi.org/10.1016/j.landusepol.2018.05.030

Tang, Z., S. Guan, M. Zhao, L. Wang, Y. Liu, and Y. Fan. 2022. Grassland transfer and its income effect: evidence from pastoral areas of the Qinghai-Tibet Plateau. Land 11(11):2020. https://doi.org/10.3390/land11112020

Tao, Y., and R. Gao. 1992. The change of rangeland and pastoral industry’s flourish-decay in history. Chinese Agricultural History 3:59-65. [Translated from the Chinese.]

The Team of Comprehensive Scientific Expedition on the Qinghai-Tibet Plateau. 1992. Rangland of Xizang (Tibet). Science Press, Beijing, China. [Translated from the Chinese.]

Weber, J. G., N. Key, and E. O’Donoghue. 2016. Does federal crop insurance make environmental externalities from agriculture worse? Journal of the Association of Environmental and Resource Economists 3(3):707-742. https://doi.org/10.1086/687549

Yang, T. T., S. Chen, Q. Y. Mu, X. Y. Liu, W. H. Liu, and T. Ye. 2021. Willingness-to-pay for yak snow disaster weather index insurance: a case study of Yushu Tibetan Autonomous Prefecture, Qinghai Province. Research of Agricultural Modernization 42(4):745-754. [Translated from the Chinese.]

Yang, T., J. Dong, L. Huang, Y. Li, H. Yan, J. Zhai, J. Wang, Z. Jin, and G. Zhang. 2023. A large forage gap in forage availability in traditional pastoral regions in China. Fundamental Research 3:188-200. https://doi.org/10.1016/j.fmre.2023.01.003

Yeh, E. T., L. H. Samberg, Gaerrang, E. Volkmar, and R. B. Harris. 2017. Pastoralist decision-making on the Tibetan Plateau. Human Ecology 45:333-343. https://doi.org/10.1007/s10745-017-9891-8

Zhang, L., G. Zhou, Y. Ji, and Y. F. Bai. 2016. Spatiotemporal dynamic simulation of grassland carbon storage in China. Science China Earth Science 59:1946-1958. https://doi.org/10.1007/s11430-015-5599-4

Zhang, Q., H. L. Xue, X. Lan, L. W. Dai, B. J. Wang, F. Q. Cui, and H. P. Tang. 2022. Livelihood vulnerability of pastoral households in the semiarid grasslands of northern China: measurement and determinants. Ecological Indicator 140:109020. https://doi.org/10.1016/j.ecolind.2022.109020

Zhang, Y., Y. Huang, F. Zhang, and Z. Tang. 2024. Effects of formal credit on pastoral household expense: evidence from the Qinghai-Tibet Plateau of China. Journal of Integrative Agriculture 23(5):1774-1785. https://doi.org/10.1016/j.jia.2023.08.013

Zhao, X., L. Zhao, Q. Li, H. Chen, H. Zhou, S. Xu, Q. Dong, G. Wu, and Y. He. 2018. Using balance of seasonal herbage supply and demand to inform sustainable grassland management on the Qinghai-Tibetan Plateau. Frontiers of Agricultural Science and Engineering 5(1):1-8. https://doi.org/10.15302/J-FASE-2018203

Zou, B., Z. Ren, A. K. Mishra, and S. Hirsch. 2022. The role of agricultural insurance in boosting agricultural output: an aggregate analysis from Chinese provinces. Agribusiness 38(4):923-945. https://doi.org/10.1002/agr.21750