Investigating the Impact of Research and Development Costs, Rule of Law and Drought on Agricultural Growth Rate in Iran

Introduction The agriculture sector is one of the most important sectors of the Iranian economy that various factors affect its growth. Drought is one of the natural factors affecting this sector. Nowadays, knowledge is the main factor of growth, production and solving different problems in different countries. Research and development (R&D) costs and the rule of law index are component of the knowledge economy. The purpose of this study was to investigate the impact of drought, R&D costs and rule of law on the growth rate of Iran's agriculture sector. For this purpose, the casual Bayesian network method was used and data were collected for the period 2002-2017. Materials and Method In this study, casual Bayesian network method was used to investigate the effect of different variables on Iran's agriculture growth. The Bayesian network provides a framework for showing the uncertainty of variables in the network, such as the effects of variables not included in the network. A common way to overcome the uncertainty of variables in the network is to measure the confidence level of variable on the condition of its parent value. The process of measuring confidence is the same as determining the prior probability of a variable given the value of the parent in the network. The Bayesian network consists of two stages, qualitative and quantitative. In the qualitative stage, the graphical structure of the network is designed, which includes directional and non-rotational connections and expresses the dependent relationships between the variables. Directional connections are arrows pointing in a certain direction, and the non-rotational connections are meant to be a way to start from one point and pass through a series of directional connections and back to the starting point of existence. Then, in the quantitative stage, the dependent relationships between the variables are expressed as conditional probability distributions. The most important point to determine the elasticity of the connections on the network, is to determine the probability distribution. In this study, variables such as agricultural growth rate, water productivity in agriculture, drought, water use in agriculture, carbon dioxide emissions, R&D costs, rule of law index and population were considered to construct causal network. One-way sensitivity analysis was also performed on the causal network. Results and discussion The results showed that increasing growth rate of R&D costs and the rule of law index and placing Iran in wet conditions increase the water productivity in agriculture and growth of this sector. Increasing R&D costs showed the best performance in improving the drought index. With higher growth rate of R&D costs (high probability, by 100%), the highest probability of drought index with the probability of 24.2% was in the high status. The rule of law scenario showed the best performance in reducing water use in agriculture. So, with higher growth rate of the rule of law index (high probability, by 100%), the highest probability of growth rate of water use in agriculture with the probability of 50.5% was in the low status. Therefore, increasing R&D costs and targeting them, as well as enforcing laws and policies in various fields such as water resources can help to increase water productivity in agriculture sector and growth of this sector.   Suggestion Therefore, in order to improve drought, it is recommended to increase research and development costs in various fields such as water. The government also proposes new policies aimed at reducing water use in the agriculture sector, increasing production, thereby increasing the income of farmers and increasing the productivity of inputs and thus increasing the profits of farmers. As a result of drought-related problems and decreasing water use in the agriculture sector, the growth of the agriculture sector is also increasing. JEL Classification: Q16, C11, Q54.