Home / Insurance

Introduction to Research Problem


CHAPTER 1: INTRODUCTION TO THE RESEARCH PROBLEM


1.1. Reasons for choosing the topic


Maybe you are interested!

Coffee is one of the main agricultural export products of Vietnam, creating nearly 2.5 million jobs, contributing 2% of the country's GDP, making an important contribution to the political - social - security stability of the country in general and the Central Highlands in particular. In recent years, coffee has made great strides, making our country the largest coffee exporter in Asia and the second largest in the world. According to the General Department of Vietnam Customs, in the 2016-2017 crop year, the country exported 1.483 million tons of coffee and the export turnover was over 3.4 billion USD. By 2017, Dak Lak province had 203,737 hectares of coffee, accounting for 33% of the country's coffee area; of which 191,483 hectares of coffee area were for business purposes; the average coffee yield reached 23.36 quintals/ha, down 26 kg/ha compared to the previous crop year; Coffee output was 447,810 tons, down 14,810 tons compared to the previous crop. The province's coffee export turnover reached approximately 450 million USD, up 24.9% compared to the previous crop (accounting for 13.3% of Vietnam's coffee export turnover, over 90% of Dak Lak's coffee export turnover) (Dak Lak Provincial People's Committee, 2017).

With its contribution to GDP, job creation and high export turnover for many consecutive years as above, in the coming years, coffee trees will continue to play a very important role in the economy of Dak Lak province and the whole country. However, coffee production and trading always contain many risks that need to be overcome. According to the 2016 report of the Vietnam Coffee - Cocoa Association, coffee trees have continuously failed, causing great damage to farming households. Especially in the 2015-2016 crop year, coffee tree productivity decreased by 20% compared to the plan. The main reason is that Dak Lak suffered from severe drought, leading to nearly 48 thousand hectares of coffee dying (General Statistics Office of Vietnam, 2016). From December 2016 to now, due to the impact of erratic rains, the entire coffee area has flowered, but the flowering is uneven. Even in many areas where coffee berries are ripening and being harvested, the trees are still flowering, causing many difficulties for farmers and reducing the productivity and quality of the next crop of coffee. Heavy rain

Introduction to Research Problem


prolonged at the time when coffee is ripe, causing the fruit to rot and fall. In 2017, the Central Highlands provinces received 16 storms, a record high, of which storm No. 12 had extremely strong winds, causing damage to people and property as well as affecting the growth of coffee trees. Many fresh and dried fruit processing factories, coffee tree purchasing agents, and storage warehouses had their roofs blown off and were severely damaged, leading to contract cancellations or price pressures on farmers. Furthermore, the increase in climate change can expand the geographic range of some insect pests. For example, La Roya coffee rust has attacked coffee factories in Central and South America at higher altitudes as the climate warms (Oxfam, 2013). As temperatures rise, crops in the growing stage can be more susceptible to insect pests during the growing season. Increased rainfall is likely to increase fungal and bacterial pathogens (M. Parry, 1990). Similar damage is occurring, with coffee berry borers becoming more prevalent in East Africa due to warming (Jaramillo, 2011). Some pests, including aphids and weevils, respond positively to higher atmospheric CO2 levels (Newman, 2004; Staley and Johnson, 2008). Climate change therefore threatens pest control and disease invasions, including insects, plant diseases and invasive weeds that pose biological risks (insect risks, pest risks, etc.) to coffee.

The above results show that farming households face many risks in coffee production and agriculture in the Central Highlands of Vietnam is under pressure due to the impacts of climate change (Jeremy Haggar et al., 2011; D'haeze et al., 2017). In particular, weather risks are an important and unpredictable issue and it has a great impact on their livelihoods. Farming households and communities always have many incentives to develop and improve strategies to cope with and manage weather risks (Bibek Acharya, 2014). Weather risk management strategies for households include improving coffee varieties, farming techniques, lending, savings, and especially insurance (World Bank, 2004; World Bank 2010; World Bank, 2015a). In the event of a weather shock, insurance is designed to protect against income loss. This allows households to avoid selling their livelihood assets.


or withdraw savings. Insurance can help farmers access new opportunities by improving their ability to borrow. In doing so, farmers will gain more secure profits, without the economic distress that leads to the loss of the ability to reproduce production (Barnett et al., 2008; World Bank, 2006). Insurance helps farmers manage risks and improve their production and welfare (Karlan et al., 2010; De Nicola et al., 2012; Radermacher et al., 2014). On the other hand, households without risk transfer mechanisms are more likely to be pushed into permanent poverty (Barrett and McPeak, 2005; Barrett and Swallow, 2006; Carter and Barrett, 2006; Carter et al., 2007).

As proposed by the Public Private Partnerships (PPP), the Vietnamese Government implemented a pilot agricultural insurance program from 2011 to 2013 but did not apply it to coffee trees. Most recently, in April 2018, coffee trees were included in the list of insured objects supported under Article 18, Chapter III, Decree No. 58/2018/ND-CP of the Government on agricultural insurance. However, by June 2019, according to Article 2, Chapter I, Decision No. 22/2019/QD-TTg of the Prime Minister on implementing the agricultural insurance support policy, it only applied to organizations and households growing rice (Appendix 6). This is a huge loss for the coffee industry. Seeing the importance of coffee yield insurance, Bao Minh Insurance Company also piloted rainfall-based coffee insurance in the 2011-2012 crop year in Dak Lak province, with about 60 households participating in agricultural insurance with an area of ​​nearly 50 hectares, with a total insurance revenue of nearly 122 million. With the above figures, it is equivalent to about 0.025 of the coffee growing area in Dak Lak being insured. A very modest number for insurance participation and can be considered as unsuccessful with this insurance product. Because, the insurance company's products have not provided practical support to help farmers limit risks.

Up to now, there have not been many in-depth studies on risks for coffee production. There are only a few studies examining risk mitigation in agriculture, conducted by domestic and foreign researchers to provide more information for


policy makers, banking industry, insurance industry and especially farmers to have specific solutions to minimize risks in agricultural production. Studies have shown factors affecting the willingness of farmers to participate in agricultural insurance such as: Research on crop insurance participation of bean and corn farmers in the United States (Sherrick.BJ et al., 2004). Research by Sarris et al. (2006) studied the willingness to participate and pay for weather insurance according to the rainfall index in Tanzania. Kong et al. (2011) studied the willingness of farmers to participate in weather insurance in Shaanxi and Gansu, China. Research by Aidoo, R., Mensah et al. (2014) assessed the willingness of farmers to participate in crop insurance. Danso-Abbeam et al. (2014) analyzed the willingness to participate in cocoa price insurance in Ghana. Abraham Falola et al. (2014) studied the willingness to participate in cocoa crop insurance among Ondo State farmers, Nigeria. Arshad et al. (2015) studied whether crop insurance is an acceptable tool for flood and drought events in rural Pakistan. Yakubu BalmaIssaka et al. (2016) studied the willingness to participate in drought index crop insurance among maize farmers in the Northern Region of Ghana. Elvis Dartey Okoffo et al. (2016) assessed the willingness to access cocoa crop insurance in Ghana. Fonta et al. (2018) studied the willingness of farmers to participate in household crop insurance in Southwestern Burkina Faso. Yanuarti et al. (2019) studied the willingness to participate in rice crop insurance of Indonesian farmers. And some domestic empirical studies: Research by Nguyen Quoc Nghi et al. (2013) assessed the need to participate in dragon fruit crop insurance of Cho Gao, Tien Giang farmers. Pham Le Thong (2013), Luong Thi Ngoc Ha (2014), Nguyen Duy Chinh et al. (2016) conducted studies on the willingness to participate in rice crop insurance in Vietnam. However, these studies all have certain limitations such as lack of in-depth research or failure to use factors including: the impact of


weather risk factors, biological risk, economic risk, labor risk to crops in general and coffee in particular (Ray.PK, 2001; Nguyen Ngoc Thang, 2017).

From the above situation, the author finds that studying the model of farmers’ willingness to participate in coffee crop insurance based on the yield index is an important and urgent issue. In order to have a truly “good” model or one that best fits the data, researchers must address the uncertainty of the model. However, using (Classical) Frequency statistical methods, empirical studies often use a single standard model in which a set of explanatory variables are regressed on an outcome variable (Raftery, 1995). In order to demonstrate the certainty, researchers then present several variations of the standard model and then make inferences or conclusions based on this single model without acknowledging the potential problem of model uncertainty (Zeugner, 2011; Moral-Benito, 2015). The empirical literature on willingness to participate in crop insurance is no exception. Therefore, when conducting empirical research, it is necessary to consider all models to select important factors and come up with the model that best fits the data (Jennifer A. Hoeting, 1999).

Faced with this problem, the author found a new and quite popular statistical method today, the Bayesian Model Averaging (BMA) method of the Bayesian statistical school. This method will allow researchers to closely examine all models. More specifically, the BMA method provides a powerful statistical selection criterion (posterior probability) to identify important factors in the model (Ali, A., and Ali, SI, 2020). In recent years, there have been many studies on the BMA statistical method in most fields of science, energy, economic forecasting, medicine, agriculture, analysis of social relationships,... but there has been no study of the BMA method on insurance (Maltritz, D. and Molchanov, A., 2013; Zhanga and Yang, 2015; Georg Man, 2015; Notaro et al., 2016; Huang, X. et al., 2017; Vuong Minh Giang and Nguyen Thanh Thien, 2017; SL Klijn et al., 2019; Yanlai Zhou et al., 2020; Ali, A., & Ali, SI,2020; Gernát, P. et al., 2020).


Therefore, in this thesis, the author will study the BMA method to apply to the model of farmers willing to participate in coffee crop insurance according to the productivity index.

Furthermore, Bayesian statistics also provides us with a regression method that analyzes dependent factors as binary variables (Millar J et al., 2018; Spyroglou, Ioannis et al., 2018; Workie MS and Belay DB, 2019; Arreola EV et al., 2020). This is also a new regression method and has not been applied in domestic and foreign empirical studies on farmers' willingness to participate in crop insurance.

Based on an overview of related documents, the author proposes the research topic "Research on the application of Bayesian statistics to analyze the willingness to participate in coffee crop insurance according to the productivity index of farming households in Dak Lak province" as a research topic for his thesis .

1.2. Objectives

1.2.1. Research objectives

Select a research model for the topic using the Bayesian averaging method (BMA).

Estimate parameters in models with binary dependent variables using frequency statistics and Bayesian statistics for comparison. Identify new points in the research model through testing.

Online construction of the research model.

To quickly predict the ability to participate in insurance of coffee growing households.

Proposing some policy implications on coffee tree insurance based on energy index

rate


1.2.2. Research questions

Which models will be selected using the BMA method? Which new variables will be selected among the statistically significant variables of those models?

How is estimating parameters in a model with a binary dependent variable different between frequency statistics and Bayesian statistics?


Building an online nomogram with a research model of farmers' willingness to participate in coffee crop insurance according to the productivity index will help quickly predict the willingness of farmers in Dak Lak province to participate in insurance.

What policy suggestions are there to increase the number of farming households in Dak Lak province willing to participate in coffee crop insurance based on the yield index?

1.3. Research object and scope

1.3.1. Research subjects

Study of Bayesian averaging (BMA) model of Bayesian statistics school.

Factors affecting the willingness to participate in coffee crop insurance according to productivity index of farming households in Dak Lak province.

The survey subjects were coffee farming households in Dak Lak province.


1.3.2. Scope of research of the topic

Bayesian mean squared (BMA) model is used in the model with binary dependent variable.

The thesis focuses on studying the factors affecting the willingness to participate in coffee crop insurance according to the productivity index of farming households in Dak Lak province.

The research time was the 2016-2017 coffee crop year.


1.4. Research method


The thesis was researched by the author using a mixed method by simultaneously using qualitative research combined with quantitative research (Creswell, 2014).

In qualitative research, through group discussions, one-on-one discussions, scientific seminars including scientists, experts in the fields of insurance, banking, coffee experts, and coffee farmers to adjust and supplement factors related to the research problem.


Quantitative research with data collected from information samples from 500 coffee farming households, resulting in a sample of 480 observations. This data source is conducted by the Dak Lak Provincial Statistics Office. The collected data will be analyzed using R statistical software.

The author uses the BMA method according to logistic regression and the BMA method according to probit regression to select factors affecting the willingness to participate in coffee crop insurance according to the productivity index of farmers in Dak Lak province. Then, the author uses the method of splitting the data into two parts: training and testing to estimate the model. Finally, the author will analyze the impact of the factors using frequency logistic regression and Bayesian logistic regression.


1.5. New contributions of the topic

1.5.1. New contributions to theory

The thesis reveals the advantages of the BMA method according to logistic regression and the BMA method according to probit regression in choosing the most suitable model for the data. Splitting the random data into two parts, one part for estimation and the other part for testing will give better and more reliable results in this study.

The thesis uses frequency logistic regression and Bayesian logistic regression to reveal the level of influence of factors affecting the willingness to participate in coffee crop insurance according to the productivity index of farming households in Dak Lak province.

Building a specific model to study farmers' willingness to participate in coffee crop insurance based on productivity index.

1.5.2. Contribution to the financial statements

The BMA method according to logistic regression and the BMA method according to probit regression have identified the research models of farmers' willingness to participate in coffee crop insurance according to the yield index that best fits the data. At the same time, the probability of impact of new factors and their level of influence on willingness to participate in coffee crop insurance are determined.

Comment


Agree Privacy Policy *