Impact of crop insurance on cocoa farmers’ income: an empirical analysis from Ghana

Risk is associated with every sector of an economy, and the pervasiveness of risk in agriculture is not new to farmers; they have, over the decades, developed ways to minimize and cope with it. The question is whether traditional strategies employed by farmers are adequate to curb unavoidable natural disasters. This study aims to see how crop insurance affects cocoa producers’ incomes in Ghana. A well-structured questionnaire was delivered to a sample of 600 cocoa farmers in Ghana’s Ashanti region, and data was collected using a multi-stage random sampling technique. Tobit and propensity score matching effect estimators were used to examine crop insurance’s impact on cocoa farmers’ income. We found that the age of a cocoa farmer has a negative effect on the farmer’s income and is statistically significant. Our result also shows that the marital status of cocoa farmers has a significant positive impact on their income. The relationship between savings and farmers’ income was positive in our estimation. It indicates that an increase in savings attitude leads to a higher income for the farmers. The result indicates that crop insurance had a significant positive impact on cocoa farmers’ income in the Ashanti region. The study recommends that the government of Ghana, with urgency, design agricultural insurance policy that can capture various farmers in the country to enhance their income and reduce poverty. Again, insurers need to promote publicity through public seminars, training, and media advertising to improve farmer awareness and knowledge of the insurance scheme.


Introduction
In Africa, more than 70% of the population makes a living out of agriculture. The tremendous contribution of the agriculture sector to Africa, particularly Ghana, is unquestionable. In Ghana, agriculture has the propensity to eradicate poverty regarded there are pragmatic policy reforms to revamp the sector. Ghana has a total land of 23,853,900 hectares (ha), out of which about 57% is arable land for agricultural production. ISSER (2013) affirmed that agriculture accounts for about 30% of the country's GDP and serves as 60% of its export. Despite the positive impacts of the sector, empirical studies disclose that numerous factors limit access to it (Magri 2007;Thaicharoen et al. 2004;Crook and Hochguertel 2005;Del-Rio and Young 2005;Akudugu 2012;Agbenyo et al. 2019).
Agriculture growth is primarily determined by the crop sector, which is highly influenced by cocoa production. Cocoa is the largest subsector of the crop sector and contributes about 30% to the agriculture sector in Ghana. Krishna (2007) noted that Ghana was one principal exporter of cocoa from 1911 to the 1970s before Cote D'Ivoire took over. Cocoa production to agriculture GDP increases from 13.7 to 18.9% between 2000 and 2006. However, Aidoo et al. (2014) emphasize that crop production is inherently risky in Ghana due to its heavy dependence on unpredictable weather factors. Ghana's agriculture is mainly small-scale and rain-fed, making them vulnerable to risks inextricably linked with the production environment. Kwadzo et al. (2013) analyzed the willingness of 120 food crop farmers in Ghana to participate in a market-based crop insurance scheme. The study shows Responsible Editor: Baojing Gu that the size of a farm, family size, and the diversification of a farm's livestock can influence a farmer's decision to adopt crop insurance. Aidoo et al. (2014) accentuate that on-farm income, farm size, education, and quantity of savings by farmers considerably affected the premium farmers vowed to pay towards a crop insurance scheme in Ghana. A study conducted by Okoffo et al. (2016) hypothesized that the larger the farm area, the more likely cocoa farmers would pay for crop insurance products. Abdulmalik et al. (2013) also hypothesized that access to credit by farmers influences the willingness to adopt and pay for crop insurance positively with the impression that access to credit will enable cocoa farmers to have a certain level of financial ease in adopting and paying for the cost crop insurance. Despite the heavily cried out literature on agricultural insurance, Nunoo and Nana Acheampong (2014) conjectured that Ghana had no formal commercial agricultural insurance scheme until 2011.
In 2010, the German International Cooperation's (GIZ) project, known as Innovative Insurance Products for the Adaptation to Climate Change (IIPACC), was initiated to support Ghana in tackling the socio-economic costs and risks associated with climate change and, in particular, the variability in rainfall patterns in Ghanaian rural communities (Stutley 2010). The policy took effect in 2011 after the establishment of the Ghana Agricultural Insurance Pool (GAIP). They execute an agricultural insurance system that is still in its infancy in Ghana. State and commercial sector entities and development partners are all involved in Ghana's farm insurance market. All public sector organizations include the National Insurance Commission (NIC), the Ministry of Food and Agriculture (MoFA), the Ministry of Finance and Economic Planning (MoFEP), the Ghana Statistical Service, and the Ghana Meteorological Agency (GMet). The NIC's primary responsibilities include approving insurance premium rates and resolving insurance claims.
Farmers' underlying preferences for three types of insurance (multiple peril crop insurance (MPCI), single-peril crop insurance (SPCI), and commodity yield crop insurance (CYCI)) were investigated and reported that many smallholder farmers encounter productivity challenges in Ghana (Owusu et al. 2021). Swollen shoot disease is a viral illness that affects a portion of Ghana's cocoa fields (Andres et al. 2018). Drought can impact production as well. These and other yield losses are covered by the MPCI (Smith and Baquet 1996;Smith and Glauber 2012). The MPCI, on the other hand, does not reimburse losses caused by low commodity prices or theft. The quantity of protection is determined by the level of yield loss coverage selected, which is based on the farmer's previous yield level (Schnitkey and Sherrick 2014). The SPCI is designed to address specific (recognized) hazards to cocoa yield loss: pests, sicknesses, droughts, and other natural disasters (Smith and Glauber 2012). Farmers are reimbursed for losses in production due to specific hazards. According to Wehnert (2018), disappointedly, the program faces two significant challenges, namely enrollment and claims. Thus, the performance of the agricultural insurance market remains low (very few products). Mensah et al. (2017) claimed that the unavoidable threat that climate change poses to Ghana's agriculture can disrupt the rural source of income. Climate change and variability are threats to poverty reduction and food security. It also makes agriculture and forestry specifically vulnerable. However, Zhao et al. (2016) indicated that crop insurance could supply an income floor for disaster victims. In developed countries like Japan, a village whose paddy experienced total losses due to an extreme low-temperature event in the summer received compensations that amounted to 64% of income from the paddy in an average weather year. This insurance level covered production costs and offered a substantial fraction of the net profit from an average year (Yamauchi 1986;Zhao et al. 2016). Agricultural insurance is a financial strategy that allows farmers to transfer production risk to a third party by paying a premium that reflects the insurer's actual long-term cost, assuming the risks occur (Mutaqin and Usami 2019). To our knowledge, no study has been undertaken in Ghana to investigate the impact of crop insurance on cocoa farmers' income after it was implemented. Zhao et al. (2016) investigated the impact of crop insurance on farmers' income in Inner Mongolia, China. Crop insurance does not have a major impact on farmers' income, according to the authors. They do advise, however, that a few specific adjustments to the crop insurance policy could boost its favorable effects. Most scholars in China also believe that crop insurance stabilizes farmers' incomes (Xing and Huang 2007;Liang et al. 2008;Sun and Chen 2011;Nie et al. 2013). Xing and Huang (2007) use historical simulation methods to estimate the effect of six governmentsubsidized insurance products on fiscal expenditure and farmers' income, based on production and agricultural price data from 31 provinces in China from 1978 to 2000. Results show that average agricultural incomes tend to increase or be more stable with higher coverage levels, and the subsidy rate also has a significant impact on farmers' income. Liang et al. (2008) and Sun and Chen (2011) use co-integration and Granger causality tests to explore the long-term equilibrium relationship between crop insurance and farmers' incomes. They found that crop insurance Granger causes increases in farmer income. Luo et al. (2011) argue that agricultural insurance reduces the disparity in losses across farmers. Because of subsidy, the program provides an income transfer to farmers from the rest of the economy. Both crop insurance and government subsidization of premiums can increase the disposable income of the peasantry. Nie et al. (2013) conclude that crop insurance can mitigate risk, increase output, smooth consumption, and fight poverty. Severini et al. (2017) investigate the impact of agricultural policy on income and revenue risks on Italian farms, as well as the implications for risk management measures. Over the period 2003-2012, balanced farm-level panel data was used to construct coefficients of variation. They discovered that direct transfers always had a significant income-increasing effect. They claim that making payments in a row reduces the risk that farmers face, allowing them to engage in riskier activities. Fallah et al. (2012) discovered that income enhanced farmer engagement in farming insurance in Iran. Thus, farmers with high incomes have much more resources to invest in new interventions contrasted to their counterparts with much less income. Sargazi et al. (2013), besides, found that farmers with higher incomes tend to take part in agricultural insurance in Iran to secure their farm products. Falola et al. (2013) examine factors that affect the willingness of cocoa farmers to obtain crop insurance. The size and crop area of the family of the cocoa farmer were positively influenced by the amount of money they were willing to pay for crop insurance. Also, McKinley et al. (2016) investigated the feasibility of a crop insurance program in Ghana.

Review on crop insurance and farmers' income
Follow-up data revealed that producers who followed certain practices had higher estimated yieldings and a lower risk of having a yield loss that is greater than 25%. These results also showed that the cost of implementing these practices was lower than expected. A recent study by Owusu et al. (2021) on crop insurance solutions for cocoa growers in southern Ghana shows that premium, payout duration, and form of payment influence product selection. The study also found that farmers who lack credit are less ready to pay for crop insurance. The study indicated that credit limitations had a significant impact on farmers' insurance participation decisions. The gap identified in this literature is that there is no current study on the impact of crop insurance on farmers' income in Ghana.

Study area and data
Ashanti Region is one of the regions that benefited from the crop insurance program, and farmers have been compensated, making the region a perfect one for this study (Fig. 1). The region currently has a population size of 4,780,380, making it the highest populated region in Ghana. Out of which, 2,316,052 and 2,464,328 represent males and The Ashanti Region has Kumasi as its capital town and lies centrally within the country's middle belt. The region falls between longitudes 0.15 W and 2.25 W, and latitudes 5.50 N and 7.46 N (Gyasi-Agyei et al. 2014). In terms of landmass, the region covers 24,389 km 2 , representing 10.2% of the total land area of Ghana. The Ministry of Food and Agriculture (MoFA) specified that 58% of this region's occupants engage in farming, fishing, and animal husbandry (MoFA 2011). The region is also noted for other crop productions such as maize, cassava, plantain, yam, cocoyam, rice, coconut, rubber, oil palm, and coffee. Owing to the brief history of the Ashanti Region, it is therefore considered suitable for the study area. Below is the map of the Ashanti Region as a study area.

Data collection
Our data were obtained from a survey administered to cocoa farmers in Ashanti Region, Ghana, in 2019. The study engages three districts, namely, Bosomtwe, Sekyere East, and Akim North District. We randomly select six cocoa-producing villages out of the three districts, including Nkowii, Pipie, Attakrom, Abono, Agogo, and Juasan. The researchers conducted a pre-test survey with eight farmers to gather data on their socioeconomic and demographic characteristics, agricultural risks, and income sources. Other areas covered included assessing the effectiveness of various agricultural insurance policies and farming systems. The objective of the pre-test was to minimize the number of redundant questions and improve the quality of the interviews. The study components were modified to include new aspects and methods based on the pre-test survey results. The study also hired three graduates to help in interviewing the cocoa farmers. They were trained on the critical questions that the farmers would need to answer. Since it is challenging to return for follow-up interviews, we prefer to collect data by interviewing at least 10 to 15 individuals in a day. This method worked for the 600 questionnaires. The district distribution of the samples is presented in Table 1 and the variable description in Table 2. The questionnaire comprises household characteristics, farm characteristics, income sources, income level, and income consumption.

Propensity score matching model
The study considered cocoa farmers' who purchased crop insurance as the treatment group, and those who did not purchase are referred to as the control group. In order to have a thorough investigation, we estimate the descriptive statistic in three main categories. Thus, we consider the entire sample, the treatment, and the control group. In addition to providing basic descriptive statistics for each variable  employed in the study, we compare the means of farmers who purchased crop insurance against those who did not. According to Rosenbaum and Rubin (1983), PSM enables the correction for selection bias concerning observable characteristics that may affect policy intervention participants. According to Ashimwe (2016), impact valuation studies grieve from three major interrelated problems with significant implications for empirical outcomes. Simtowe et al. (2012) indicate that the first problem is the causal effect between treatment group and their impact on the outcome. The second problem is the omission of confounding covariates, while the final problem is purported to be counterfactual. Caliendo and Kopeinig (2008) posit that the PSM method does not solve the problem as it is perceived, hence the need to estimate the sensitivity test to investigate whether essential variables were overlooked in the evaluation and sensitivity of estimated treatment effects in the presence of unobserved heterogeneity.
Matching is an essential estimation for treatment effect to remove over bias and estimate the treatment effect using observational data (Baser 2006). The goal of propensity score estimation is to balance the covariate distribution in the treated and control groups (Rosenbaum and Rubin 1983). Propensity score matching employs a predicted probability of the treated and control grouped in this case, cocoa farmers who purchased crop insurance and did not purchase crop insurance, respectively. To maintain consistent matching between the results, four different techniques for the estimation of the propensity score matching were employed, namely, nearest neighbor (NNM), radius (RM), kernel (KM), and local linear regression matching (LLRM). With the NNM method, the study seeks to order treated and control groups randomly, and then select the first cocoa farmers who purchased crop insurance and find one cocoa farmer's who did not purchase crop insurance with the closest propensity score (LaLonde 1986;Baser 2006).
We employed the definition of Rosenbaum and Rubin (1983) as the probability of being part of the treatment given pre-treatment characteristics as: where FI = {0, 1} is a dichotomous variable representing whether a cocoa farmer purchased crop insurance in 2011 where yes represents 1 and 0 otherwise, is the multidimensional vector of pre-treatment characteristics of a farmer, and P( ) is the propensity score. We estimate the impact of crop insurance on cocoa farmers' income and the average treatment effect (ATT) on cocoa farmers' who purchased crop insurance after matching was deduced. Baker (2000) and Ashimwe (2016) disclosed that the expected value of ATT is inferred as the change between expected outcome values (income of farmers) with and without treatment for farmers who purchased crop insurance in the treatment group.
(1) P( ) ≡ Pr{FI = 1| } = E{FI| } The initial step for estimating PSM is the binary estimation of factors anticipated to influence farmers' income after purchasing crop insurance. The Tobit regression model was employed to analyze the impact of crop insurance on cocoa farmers' income. The valuable econometric model for analyzing factors affecting cocoa farmers' income is the Tobit model shown in Eq. (3). The focus is on the distribution of ( IF 1i | i = 1 ); i represents the farmers in this equation, 1 i and 0 i are the predictable outcomes in the two counterfactual situations of farmers who purchased and farmers who did not purchase crop insurance, and ϑ i represents the treated group. The Tobit model presumes a latent unobserved variable Z * i that depends linearly on Υ i through a constraint vector . There is a generally distribute error term i to capture the random influence on this association. The Tobit model can be defined as: where Z * i is a latent variable (monthly average income of cocoa farmers) In the Tobit model, the explanatory variables are a function of latent variables defined by observable household and agricultural factors (both exogenous and endogenous factors), as well as the error term. The empirical model is shown below.

Descriptive results
Evidence from the fundamental descriptive analysis indicates the systematic differences between treatment and control groups. From the descriptive in Table 3, we can deduce that income, age, marital status, education, lnINC * i = * 0 + * 1 Age i + * 2 Edu i + * 3 Gen i + * 4 AgeF i + * 5 Fsz i + * 6 AC i + * 7 Hsz i + * 8 Fexp i + * 9 MS i + * 10 S i + * 11 OTI i + * 12 EO i + i awareness of crop insurance, age of farm, extension officer, and savings recorded a significant difference between treatment and control groups. In contrast, other variables did not exhibit any significant difference between the two groups. However, a focus on our interest variable (income) implies that farmers who had the opportunity to participate or purchase crop insurance improved their income compared to farmers who did not purchase crop insurance. This finding is consistent with Zhao et al. (2016). They identified a difference among farmers who purchase crop insurance to have an average income than farmers who do not pay for it. They, however, likewise disapproved of the findings on the basis that the DiD estimator is biased.
They equally attribute the increment in farmers' income to be nominal since it is not inflationary adjusted as in the case of this study. This finding cannot be reliable since it has been confirmed that there is an imbalance among the distributions. Hence, the study needs to employ the propensity score matching method to reduce the bias identified among the treated and control groups based on the observable covariates and better compare the groups.

Factors that influence cocoa farmers' income
Multi-collinearity occurs when the correlation between the independent variables in a regression model is high. Modeling and analysis can be complicated when there is a high degree of correlation between variables. Prior to data analysis, the contingency coefficient test was conducted to diagnose colinearity and exclude independent variables substantially associated with each other, as shown in Table 4. The weak coefficients between the independent variables confirm there is no multi-collinearity among the variables employed in our study. The results of Tobit regression are presented in Table 5. Our result indicated that farmers' age coefficient had a negative effect on their income and was statistically significant. The age of a cocoa farmer negatively influences the willingness to pay for crop insurance. An additional year in the age of a cocoa farmer will negatively influence his or her income. The probability is − 0.53, all other things being equal. The result is consistent with previous studies such as Falola et al. (2013), Wairimu et al. (2016), Okoffo et al. (2016), and Langyintuo and Mekuria (2005). The marital status of cocoa farmers had a significant positive impact on their income, as expected. This estimate is statistically significant at the 10% level. The possible interpretation of this outcome is that married farmers may have their spouses engaging in different work to reduce the burden on their family income. Danso-Abbeam et al. (2014) conjecture that married farmers have responsibilities and compel them to engage in activities that reduce their vulnerability to risks. Access to credit had a significant negative impact on farmers' income and was statistically significant at the 10% level. Not surprising as most of the respondents in the study area indicate that they lack access to credit. Access to credit is a significant problem for most African farmers as financial institutions find it risky to loan farmers. This result implies that the non-accessibility of credit leads to a decrease in farmers' income by a probability of − 0.021, all things being equal. Savings had a positive relationship with farmers' income. It implies that a unit increase in the savings attitude of farmers leads to an increase in farmers' income by 26.4%. We found that marital status was significant and positively correlated. It is likely that cocoa farmers who are married have a higher probability of adopting crop insurance and will be willing to pay since they have more responsibilities and would want to reduce the family's vulnerability to risk. The result is in line with Danso-Abbeam et al. (2014). Farm age has a negative effect on farm income. The possible explanation for the negative coefficient could be associated with the loss of fertility of the soil, which could impact total output and hence negatively influence their income. However, statistically, it is insignificant in our study. Similarly, farm experience has an insignificant negative influence on farm income in this study. The study revealed that the longer a farmer has been in a farming business, the more likely they avoid buying an insurance policy. This is because they could be prone to experiencing weather-related risks. In addition, there is the possibility that a lack of understanding of crop insurance is to blame for a negative coefficient. Danso-Abbeam et al. (2014) also argue that an experienced farmer will show less interest in adopting and paying for crop insurance since he would like to use his traditional experience to adopt different strategies in controlling risk.  farmers who purchased crop insurance from the year of its implementation and cocoa farmers who did not purchase crop insurance since its implementation. Before a PSM estimation, it is necessary to check for possible violations of some conditions; a typical example is the overlap assumption. There is a need for both treated and control groups to satisfy the common support condition. Thus, according to Caliendo and Kopeinig (2008), both groups must be within the standard support region. Evidence from the visual assessment of Fig. 2 on the density distribution between the groups indicates that the treated and control groups are within the region of common support. From  Fig. 2, the upper region in red implies the distribution of the treated group, whereas the bottom in blue represents the control group. The y-axis denotes the propensity scores for the two groups. By implication, each cocoa farmer had a positive probability of being a buyer of crop insurance or not a buyer of crop insurance. It validates the common support assumption that necessitates each cocoa farmer who purchased crop insurance to have a corresponding non-crop insurance buyer as a match (Austin 2011;Ashimwe 2016).

The impact of crop insurance on cocoa farmers' income
This study employed all the four techniques mentioned above to estimate the differences between the covariates of cocoa farmers. With RM estimation, each treated element was matched only with the control element whose propensity score falls in a predefined neighborhood of the propensity score of the treated unit (Dehejia and Wahba 2002;Baser 2006). According to Baser (2006), the pros of the NR method are that it uses only the number of judgment elements accessible within a predefined radius, hence allowing for the use of extra elements when suitable matches are available and fewer units when they are not. One disadvantage of this method is the decision of precise radius to use compared to KM. All treated elements are matched with a weighted average of all controls, with weights inversely proportional to the distance between the propensity scores of the farmers who purchased and did not purchase the insurance. All cocoa farmers who did not purchase crop insurance contribute to the weights achieving lower variance. According to Caliendo and Kopeinig (2008), this can be considered a plausible counterfactual.
These four methods have evidentially portrayed no systematic differences in the distribution of covariates between treated and control groups. Insignificant p-values of the likelihood ratio and a reduction in bias after matching for the covariates balance tests, according to Rosenbaum and Rubin (1985), should be used to specify the estimation. Caliendo and Kopeinig (2008) later posit that variances are predictable before matching; nevertheless, there should be a balance in the treated and control groups after matching the covariates, indicating that no significant difference is found (Fig. 3). Evidence from the current study shows no significant difference found between the treated and control groups after matching. Table 6 presents summary statistics, including the standardized mean and median bias coupled with the pseudo-R 2 . According to Rosenbaum and Rubin (1985), the standardized mean and median bias differences between the treated and control groups should not be more significant than 20%. The difference between the standardized mean and median bias when greater than 20% is considered large. Evidence from Table 6 indicates that among the four matching algorithms employed in the covariate balancing tests, radius matching, and kernel-based matching, stands out to be the appropriate matching algorithm. This assertion is based on the reduction rate of standardized mean and median bias. The mean bias before matching for all the four matching techniques is 11.5, which has reduced to 3.8, 2.2, 2.1, and 3.6 after matching for the nearest neighbor, radius, kernel-based, and local linear regression matching, respectively. The median bias also reduces from 8.5 to 3.3, 2.1, 1.9, and 3.3 for the nearest neighbor, radius, kernel-based, and local linear regression matching. Based on the results, the appropriate matching techniques we adopt are radius and kernel-based propensity matching. Considering the mean and median biases of the RM and KM are below 10% indicates a good match between cocoa farmers who purchased and did not purchase crop insurance. Again, pseudo-R 2 of the RM and KM before and after matching indicates the probability that farmers who did not purchase crop insurance are likely to purchase (Sianesi 2004).
Before matching, the pseudo-R 2 was 0.037, but it dropped to 0.001 for RM and KM, which is very low and significant enough to support the assumption that there are no variations in the distribution of variables between the treatment and control groups. As a result, this study asserts that the matching method was able to effectively balance the circulation of covariates between the treated and control groups, and there is a reliable counterfactual on the assertion that this study achieved  Fig. 2 Propensity scores distribution before matching low pseudo-R 2 values, insignificant p-values, low standardized mean bias, and high total bias reduction. These findings indicate that there is no consistent difference in the covariate distribution of income between the treatment and control groups. As a result, the crop insurance intervention in the community could result in a unit difference in farmers' income between the two groups with the possibility of increasing.
The treatment effect (ATT) of crop insurance intervention on cocoa farmers' income is presented in Table 7. Table 6 reports the treatment effects based on nearest neighbor, radius, and kernel matching algorism. The results for comparison of cocoa farmers who purchased and did not purchase crop insurance are statistically insignificant but in the anticipated positive direction for the neighbor and kernel matching estimations. However, the radius matching had a positive and significant impact on cocoa farmers' income. This implies that farmers' participation in crop insurance has a positive impact on farmers' income. The interpretation of this result would imply that a cocoa farmer who purchases crop insurance tends to earn a higher income than their counterparts who did not purchase crop insurance by 77.6% based on the radius matching technique. The result is consistent with Zhao et al. (2016) where they established that there is no significant impact of crop insurance on farmers' income in inner Mongolia in China. The estimation also agrees with Varadan and propensity scores AFTER matching treated control Kumar's (2012) study, which revealed that high use of farm inputs and production risks are absorbed by agricultural insurance. Nahvi et al. (2014) found a significant positive relationship between income and agricultural insurance in Iran. Similarly, Yanuarti et al. (2019) also found a positive impact of crop insurance on Indonesian farmers' income.

Conclusion and policy implications
Most studies on the impact of crop insurance on farmers' income focus on regional time series analysis and are carried out in developed countries where crop insurance has been adopted for decades compared to Ghana. We examine the factors influencing crop insurance and its impact on farmers' income in Ghana. Our study contributed to the scarce literature on this topic by using survey data and propensity score matching method coupled with an average treatment effect of the intervention on farmers' income. We control for endogeneity and self-selection bias by adopting a propensity score matching and average treatment regression model. The estimates show that household size, gender, age of farm, access to credit, and farm labor negatively influence farmers' income. Farmer's age and farm experience had a negative effect on income but were statistically insignificant. The marital status of cocoa farmers had a significant positive impact on farmers' income. The study conjectures that purchasing crop insurance for a cocoa farm led to an increase in farmers' income. As evidence from the findings that crop insurance has a positive impact on farmers' income, the study recommends that the government of Ghana design agricultural insurance policy that can capture a wide range of farmers in the country with a matter of urgency. Government funding usually takes the form of a direct premium subsidy to the producer. This usually is common in European nations such as the Czech Republic, France, Austria, and Slovenia. These countries have been prosperous in having a welldeveloped agricultural insurance system for their farmers. The government of Ghana should adopt such policies and intervene in the policy in the country. The findings of the study are strong enough to suggest that farmers' income increased as a result of their participation in crop insurance; as a result, practical steps are needed to promote farmers' desire to participate in insurance programs. Farmers must be educated on the need of crop insurance for farming operations, either through the insurer's aggressive marketing of the insurance program or through cooperative activities. In this sense, cooperative societies in Ghana need to be encouraged so that they may effectively advocate for their members and disseminate information.
These empirical results can serve as a valuable source of information for developing countries and their ministries.
This research can also serve as a reference for other countries and institutions developing crop insurance schemes.
Author contribution W.A.; conceptualization, data curation, formal analysis, and investigation. G. N.: literature review and data curation. Y.J.: supervision, review, and editing. All authors have read and agreed to the manuscript.

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations
Ethics approval and consent to participate N/A

Competing interests
The authors declare no competing interests.