The study considers the Ethiopian economy. It relied on secondary data (time series data) to investigate the impact of climate change on Ethiopian Agricultural GDP growth. The major sources of data for the study are the Ethiopian National Metrology Agency (NMA) and satellite data (TAMSAT) for climate variables, the National Bank of Ethiopia (NBE) for real agricultural GDP and the Central Statistical Agency (CSA), World Food and Agriculture Organization (FAO), World Bank Development Indicators Data Base, Ethiopian ministry of Agriculture for fertilizer, improved seed, agricultural land and labour force data. Depending on obtained data, the researcher used descriptive analysis to study the trend of variables over time and used the econometric model for detailed empirical analysis.
A yearly time series data on Real Agricultural Gross Domestic, Coefficient of variation of rainfall and mean monthly annual temperature (as an indicator for climate change), Land input used for agricultural production, total labour force employed, an area under fertilizer usage, and Improved seed usage is gathered covering the period from 1992/93 to 2017/18. The choice of the period is based on the availability of data. Relying on data on these variables, the econometric model is used to estimate the long-run and short-run elasticity of the variables.
In this study, to test the existence of a long-run relationship between the dependent variable AGDP and the other explanatory variables, the autoregressive distributed lag bound testing approach developed by Pesaran et al. (2001) is used. This approach is chosen since this technique has the advantage of being more efficient for studies with a small sample and applies to series that are integrated of order 1, level 0 or mutually cointegrated, unlike the traditional tests such as Engle and granger, Johansen and Juselius, etc.
Following the literature (particularly Eberhardt and Teal, 2012), a Cobb-Douglas production function can be applied to analyze the impact of climate change on agricultural production output. Thus, to quantify the impact of climate change on the economic performance of Ethiopian agriculture, this analysis uses the Autoregressive distributed lag approach on the following empirical production function, which relates agricultural output to the various factors of production.
The functional equation is based on the assumption that a country's agricultural production is a function of technology, capital, labour and climate. Based on the extant literature, a model is specified to perform the relationships between Real Agricultural Gross Domestic Product, Coefficient of variation of annual mean monthly temperature and Coefficient of variation of mean annual rainfall. Since Agricultural Production broadly is a function of capital, labour, and technology; Land input used for agricultural production, total labour force employed, fertilizer usage and improved seed usage are included in the model as control variables. That is: -
RAGDP = f (CVT, CVR, LD, FA, IS, LF)
The model employed in this study can be written as follows.
The coefficient of variation (CV) is defined as the ratio of the standard deviation to the mean. It shows the extent of variability with the mean of the population. According to Hare (2003), CV is used to classify the degree of variability as less (CV < 20), moderate (20 < CV < 30), and high (CV > 30).
Coefficient of variation as a measure of climate change-induced variability is used since it includes two commonly used measures of temporal climate variability, mean and standard deviation, in one value. The mean value has some drawbacks in providing full information about a population's true characteristics since it is highly affected by extreme values. The coefficient of variation gives an estimate of variability around the average around 67% of the time. It is a measure of relative variability.
The Rainfall and temperature data used to calculate the coefficient of variation are national-level data and annual data is used in the computation is satellite data. But to further analyze the consistency of the empirical result coefficient of variation in each synoptic station (those presenting a general view of the whole country) is computed and the average result is taken in the regression analysis. Furthermore rainfall and temperature coefficient of variation of the country's main wet season, the Meher season has been calculated and analyzed to check the robustness of the empirical result.