Agricultural Export and Economic Growth with application of Co-integration model: The Case of Live Animal, Meat and Leather Products in Ethiopia

Economic growth is an increase in productive capacity of an economy. There is different school of thought on economic growth including from the classical to solo-swan economic growth theories. Export sector of the country were seen as engine to economic growth through economies of scale, accessing the markets and expanding trade. This study was conducted in order to investigate the relationship between export and economic growth, special focusing on agricultural export (meat, live animal and leather export) in Ethiopia. The data used for this study were a time series data from 1971 to 2018. The data were analyzed using co-integration model and vector error correction model. The �nding of the study showed that consumer price index and live animal export) were affecting real GDP negatively while meat export, exchange rate and gross �xed capital formation affects real GDP positivity in the long run. The result from the short run matrix showed that the coe�cient of error correction term for the equation is signi�cant and negative, which indicates there is a reasonable adjustment towards the long rung steady state. This guarantees that the real GDP may temporally deviate from its long equilibrium value and converge to its equilibrium. The value of error correction term (-0.31146) indicated that about 31% of the deviation of real GDP from its equilibrium value is eliminated every year, as a result the full adjustment would require a period of more than three years. Live animal export is signi�cant and its coe�cient is negatively related to real GDP, indicating that it has unfavorable effect on real GDP in short run. Export of meat, Exchange rate, total labour, and GFCF were found is signi�cant and its coe�cient is positively related to real GDP in the short run. The result from the causality test revealed that there is bi-directional causality between meat export and real GDP and unidirectional causality between live animal, leather export and real GDP.


Economic Growth and International Trade
Economic growth is an increase in productive capacity of an economy which is desirable to bring economic development. The importance and component of economic growth has been centered developed and theorized in different school of thoughts. The rst school of thought, classical theory of economic growth, explains economic growth based on the class structure of the capitalist economy and identi es three classes: workers, capitalists and land owners which have their own speci c role in the economic process. This theory assumes the rate of growth of the economy is determined by the interplay between savings and population growth rate. In general, the Classical economists suggested a number of factors that promote economic growth in explaining the theories of economic growth. They started from the basic factors of production such as labor, capital, land and technology and continued to the non-economic factors such as political stability, the security of private property, the role of laws and institutions, the expansion of towns and growth of population; and non-market variables such as education and customs. Most of these variables are also considered as the factors of production and are used to explain growth in the positively to each other in the course of the Chinese economic development. Chang et al (2000) used multivariate causality analysis incorporating imports as a factor in the relationship between exports and output in the case of Taiwan and found no support for the export-led growth hypothesis during the period of rapid growth in Taiwan . Keong, Yusop and Liew (2003) run vector autoregressive model and multivariate co-integration for Malaysia and found a long-run association between the variables considered. The results of the error correction model revealed that all variables except exchange rate Granger cause economic growth in the short run. This led them to con rm the validity of the export-led growth hypothesis in the case of Malaysia both in the short -and long -term. The result further suggests that the growth rate of capital formation and imports have positive impacts on economic growth, while labour has a negative impact in the short run. Sharma and Panagiotidis (2004) test the export-led growth hypothesis for the case of India using different approach and the results strengthen the arguments against the export-led growth hypothesis for the case of India.

Data Source And Collection Methods
This study was carried out in Ethiopia on three selected export commodities namely live animal, meat and leather products. The study is based on secondary data from National Bank of Ethiopia, Ministry of Finance and Economic Development, Ethiopian Revenue and Customs Authority, Export Promotion Agency and the Ethiopian Economic Association statistical data base covering the period from 1971 to 2018.

De nition of variable
Live animal export (Animal_Export) It is the total amount of animals export without any process in the country. Live animal exports have been, and remain, a contentious issue. Many countries have been made to end the live export trade, rather to replaced with processed meat and leather products. The live animal export was done with the objective of relieve the foreign exchange constraint, allowing capital goods to be imported to boost economic growth. Therefore positive relationship will be expected on economic growth.

Meat Export (Meat_Export)
Meat is being exportable food items that have been contributing to the foreign export earning and GDP growth of Ethiopia since the beginning of meat processing. Even though the country is blessed with huge livestock resources, the share of meat and other slaughter by-products exported from the overall export commodities is not more than 2%, because the meat export is a catalyst for productive growth and to generate foreign currency, it is expected to have a positive relationship between meat export and economic growth.
Leather and leather product export (leather Export) It is the total amount of leather and leather product export and it is both processed and unprocessed. It has also a potential to means for productive growth and to generate foreign currency, it is expected to have a positive relationship between leather export and economic growth.

Total population
The workforce or labour force is the labour pool in employment. It is generally used to describe those working for a single company or industry, but can also apply to a geographic region like a city, state, or country. Within country, its value can be labeled as its "Workforce in Place". The workforce of a country includes both the employed and the unemployed.

Gross xed capital formation (GFCF)
Gross xed capital formation is used as a proxy for capital. It measures the value of acquisitions of new or existing xed assets by the business sector, governments and "pure" households (excluding their unincorporated enterprises) less disposals of xed assets. GFCF is a component of the expenditure on gross domestic product (GDP), and thus shows something about how much of the new value added in the economy is invested rather than consumed.

Gross domestic product (GDP)
is de ned as the monetary value of current nal goods and services produced by factors of production located within a country within a given period, usually a year. It is consider as the dependent variable in this study.

Real exchange rate (RER)
It is the purchasing power of a currency relative to another at current exchange rates and prices. It is the ratio of the number of units of a given country's currency necessary to buy a market basket of goods in the other country, after acquiring the other country's currency in the foreign exchange market, to the number of units of the given country's currency that would be necessary to buy that market basket directly in the given country. In this view, a positive correlation between real exchange rate and economic growth is expected.

Consumer Price Index (CPI)
It is the change in the cost to the average consumer acquiring a basket of goods and services and it is considered as the proxy for in ation, as the result, it is expected to have a negative relationship with economic growth.

Model speci cation
This study was used the Solow-swan economic growth model as the starting to explain the economic growth relationship which explains capital accumulation, labor and technological progress, with simple production function; It is also possible to expand this function by including agricultural export (only livestock sector is consider for this study); Y t = f L t , K t , Meat_export t , Animal_Export t , leather_export t ……………………………2 The formulation of this equation can be amended by adding variable including exchange rate and consumer price index as a control variable; Y t = f L t , K t , Meat_export t , Animal_Export t , leather_export t , RER t , CPI t , ……………3 Finally, by taking the logarithm for discard the difference in the units of measurement for the variable and to minimize the gap between the variable, then; logY t = β 0 + β 1 logL t + β 2 logK t + β 3 logMeat_export t + β 4 logAnimal_Export t + β 5 logleather_export t + β 6 logRER t + β 7 logCPI t + u t ……. . .4 where; LogY t = natural logarithm of real gross domestic product LogL t = natural logarithm of labor force LogK t = natural logarithm of gross domestic xed capital formation LogMeat_Export t = natural logarithm meat export LogAnimal_Export t = natural logarithm live animal export LogLeather_Export t = natural logarithm leather and leather products export LogRER t = natural logarithm real exchange rate LogCPI t = natural logarithm consumer price index u t = error term β 0 = constant term and β 1 -β 7 = are the parameters of independent variables that are going to be estimated

Method of Data Analysis
This study used both descriptive and Econometric statistics for data analysis. The descriptive statistics were used to show the trend of agricultural Export. To assess the relationship between the agricultural export and economic growth, the co-integration models were used. First, the stationary of the time series data were tested by Augmented Dickey Fuller test and optimal lag length were determine. Secondly, johansen co-integration test were performed in order to determine the number of existed co-integrated vector for the variable of interest. The main purpose for this test was to nd out whether there is a long run equilibrium relationship between variable. Thirdly, vector error correction methods were preformed in order to analyze the long run and short run relationship between variable. Finally, conclusion were drawn based on the hypothesis and the results.

Trends In Meat Export
Ethiopia has great potential for meat and meat related products exports. In 2017/8, the volume of global meat produced was 308.5 million tons and Ethiopia share of export was 659305 tones. In Ethiopia cattle, goats, sheep, camel and poultry, are used as resource for meat production; however, the rst three species are the most common. Despite the fact that Ethiopia is the tenth largest livestock population in the world, the production of meat is still low and contributed for about 0.2 percent of the world total meat production, of which most is sheep and goat meat. The reason behind for this include but not limited to very low off take rates; large number of animals that by-pass abattoirs and are exported live both illegally and legally, producers who are not commercially ( ) ( ) the top exports of the country. Hides and skins export ranked second to coffee in the 1970s and early 1980s. However, the export ban imposed on hides and skins in 1986 resulted in a decline in the export volume (Abebe & Schaefer, 2013). The nding of this study indicate that the Ethiopian export of leather and leather products shows an increasing trend (Fig. 4). Ethiopia leather (specially raw hides and skins and semi-processed leather products) export were

Unit Root Testing For Stationary
It is well know that most of the macro economy data were have the non-stationary properties and they tend to have a deterministic and/or stochastic trend. In time series analysis, it is important to consider a number of issues that can potentially in uence the estimation. In order to give good reason for the economic theory, stationary properties of the variables behind the time series models shall be conducted in order to avoid spurious regression, meaning, there will not be have meaningful relationship between those variables. Thus, it is a good move to test the stationary properties of the variables for the next step.
This study have tested the stationary properties of the variables using the commonly used method of Augmented dickey fuller (ADF) test. The result from the ADF test below revealed that almost all variable have non-stationary property at level and they became stationary after rst differencing, they have ful ll and leads to apply the co-integration or long run equation test (Table 3).

Lag length Determinations
The lag length determination essential to avoid the previous year observations dependent variable or independent variable to be considered as the explanatory variable. This implies that, Since the dependent variable is time series type the previous year observation either the dependent or the independent variable may be considered an explanatory variable. To determine the lag length, there are well known methods and criteria which lead to different outcomes and needs subjective judgments. In this study, the maximum four lag of dependent as well as the independent variable considered based on AIC and HQIC criteria ( Table 2).  (Table 4). This is a proof for the long-run relationship among variable. This co-integration rank test also led for the application of the johansen method instead of the single equation based equation-based Engle-Granger two-step procedure.

Vector Error Correction Model Estimation
The co-integration rank test revealed that the data has six co-integration relationship based on the johansen co-integration test. Thus, VECM consists of two parts: the matrix of long-run co-integrating coe cients that is used to derive the long-run co-integrating relationship, and the short-run coe cients which is for the short-run analysis.

Long run co-integration
The long run relationship test tries to show the co-integration of the dependent variable and the independent ones, which emphasizes the existence of long run co-movement In the two types of variables. These imply that the study is trying to see long run rst order of integration. Based on the VECM matrix of long-run co-integrating coe cients result, out of the seven variable, ve variable are found to be statistically signi cant and the sign of two variables (consumer price index and live animal export) were affecting real GDP negatively while meat export, exchange rate and gross xed capital formation affects real GDP positivity in the long run. In order to clearly understand the above and interpret the result, we can estimate the vector error correction equilibrium relationship normalized on real GDP. The equation of this model incorporates a corrective mechanism by which previous disequilibrium in the relationship between real GDP and one or more The above equation showed the in the long run real GDP can be explained by the Meat export, animal export, Exchange Rate, CPI, total labor and GFCF.

Short run relationship
The result from the short run matrix showed that the coe cient of error correction term for the equation is signi cant and negative, which indicates there is a reasonable adjustment towards the long rung steady state. This guarantees that the real GDP may temporally deviate from its long equilibrium value and converge to its equilibrium. The value of error correction term (-0.31146) indicated that about 31% of the deviation of real GDP from its equilibrium value is eliminated every year, as a result the full adjustment would require a period of more than three years. Live animal export is signi cant and its coe cient is negatively related to real GDP, indicating that it has unfavorable effect on real GDP in short run. Export of meat, Exchange rate, total labour, and GFCF were found is signi cant and its coe cient is positively related to real GDP in the short run.

Wald Causality Test
The co-integration test can determine the existence of causality between variables it cannot determine the direction of this relation. If two variables are cointegrated then a relationship will exist that can be measured by Granger causality test. The result from the causality test revealed that there is a short run causality between meat export, live animal and real GDP. In addition, there is bi-directional causality between meat export and real GDP and unidirectional causality between live animal, leather export and real GDP.

Conclusion And Recommendations
Agricultural export has a relationship with economic growth in many aspects and it is highly recommended to diversify the export in order to fully get bene t from international trade. The meat export by Ethiopia has long as well as short run relationship with real GDP, indicating the importance of the sector, to increase the impact, effort should be done towards modernizing of the production, processing, transportation and promotion. To facilitate this, export incentives, developing the infrastructure include road net work and other efforts is recommended. speci cally; Market infrastructure and service facilities along borders, and implementation of strategies with neighboring countries for legalizing trade should be developed Major infrastructure for livestock marketing like improved slaughter house, storage and quarantine facilities at required sites should be present Sanitary conditions at slaughtering facilities along with employees training could be in a place and cold chain management: Chiller facilities, quality labeling and packaging materials at abattoirs must be su cient The live animal export has unfavorable impact on the economic growth in the long run but is has causal impact in the short run, it is unidirectional causality.
Although the government has limited direct intervention in the livestock sector, it plays a facilitation role through the provision different services and incentives. As part of its overall economic development plan, there is a need to improve foreign exchange earnings through increasing livestock export items in terms of quality and quantity. To make this a reality, it is expected to further improve its facilitation and controlling role for actors with a view to making them vibrant and garnering sizeable revenue from the sector. Leather and leather products has long run relationship with gross domestic product and hence with economic growth but did not have short run causality, this is because of competitiveness of the products in the international market, the product quality and manly concentrated on the raw material export like hide and skin. For developing competitiveness, there should be due attention in the technology transfer (knowledge, new designs, techniques and methods while bringing their design) in between local rms and foreign customers. So it is highly recommended that to work more on it because export strategy will drive the local one.  Trends of leather and leather products export