ESTIMATION OF THE IMPACT OF SOME VARIABLES OF AGRICULTURAL ECONOMIC POLICY ON THE IRAQI DOMESTIC AGRICULTURAL PRODUCT FOR THE PERIOD 1994-2015 USING THE METHOD OF COINTEGRATION AND THE ARDL MODEL

0101 ESTIMATION OF THE IMPACT OF SOME VARIABLES OF AGRICULTURAL ECONOMIC POLICY ON THE IRAQI DOMESTIC AGRICULTURAL PRODUCT FOR THE PERIOD 1994-2015 USING THE METHOD OF COINTEGRATION AND THE ARDL MODEL .R. F. Muhammed * A. D. K. Alhiyali Researcher Prof. Dept. Agricu. Economics Coll. Agric. Univer. Baghdad Rafidfattah1967@gmail.com adk_1966@yahoo.com ABSTRACT The calculations of gross domestic product (GDP) show the contribution of each sector, whether Service fully or productively, to the formation of the national income.The agricultural sector is an important sector, although it did not take the lead in the composition of GDP because of the large contribution of the oil sector, especially in recent years, which witnessed the return of Iraq to the international oil market. The research aimed to measure the impact of some economic variables in agricultural GDP and analyze the role that these variables can play on the growth of this output, which can promote growth in the Iraqi agricultural sector. The research was based on the quantitative method to arrive at its results by following one of the modern methods to study the causal relationship, the method of multivariate cointegration, the ARDL model and the test of the causal relationship to determine the direction of the relationship between the economic variables studied, based on the assumptions of the economic theory. The study found that there is a long-term effect between the agricultural GDP index and the other economic variables under study and that there is a causal relationship between the long term and the short term. Keyword: Agricultural net value added. CUSUM, Accumulation of fixed agricultural capital *Part of Ph.D. dissertation of the first author


INTRODUCTION
Agricultural production faces a high degree of risk as it requires a relatively long period of time from the use of inputs, and the intervention of many factors beyond the control of the agricultural product in determining the final output in quantity and quality and because of this nature of agricultural production, the fluctuations in the volume of production is one of the main features In general, and in Iraq in particular, the facts indicate that Iraq was influenced by the political, economic and legislative variables witnessed by the length of the study period and the effect was therefore reflected on the effectiveness of its contribution to GDP.The economic literature is rich in applied studies that support the positive impact of some of these variables on the growth of agricultural production.In 2016, was an econometrical study of the effect of government subsidies on the growth of agricultural production in Algeria was carried out using the self-regression model of lag times using annual data for the period (1970-2011) (14), saying that there is a negative impact to support agricultural inputs on agricultural output in the long run.Also was studied the impact of CAP subsidies on total agricultural productivity (TFP) in the EU (EU) (24), the benefits have a negative impact on the productivity of the farm and after the separation of these subsidies has become a positive impact on productivity in many countries of the Union.In addition to the above, the subject of price policy and its partial and total effects has been studied by (1,2,3,5,6,8,11,15,16,19,20,26,27).The problem of the research is that the agricultural policy of the state affects the growth of agricultural GDP by controlling a range of agricultural economic variables such as: (amounts of subsidies to the agricultural sector, agricultural loans, and quantities received from strategic crops such as wheat, rice, barley, this effect is transferred to the agricultural market either directly through short-term impact or indirectly through the long-term impact.In light of this relationship, which can arise between these variables and the total agricultural output, the question arises about the nature of the relationship between short and long term.The research assumes that there is a causal relationship between some agricultural economic variables and the agricultural local product and the impact of this relationship in the short and long term.The aim of the research is to determine the impact of some variables of agricultural economic policy on the agricultural GDP.

MATERIALS AND METHODS
The research was based on the quantitative method of reaching its results by following one of the modern methods to study the causal relationship, the method of multivariate joint integration, the ARDL model and the test of the causal relationship to determine the direction of the relationship between the economic variables studied, based on the assumptions of economic theory.In addition, the research data were based on the World Bank, the Central Bureau of Statistics, the National Accounts, the Food and Agriculture Organization of the United Nations (Statistical Yearbooks), some research papers, university papers and the Internet was also relied on.The model was generally formulated as follows: Where: In order to test the existence of cointegration between the variables in the model, the hypotheses are formulated as follows: Null hypothesis: There is no cointegration Alternative hypothesis: existence of cointegration

STATISTICAL RESULTS
The first step is to examine the time series stability grades.This was done by the software developed by the (almusabah) and with ADF and PP, it is important to note that testing the stability of variables is not a necessary condition to start the application of the ARDL model, but the model does not work accurately if there are some variables stable in the case of the second difference, which has been confirmed by the fact that the variables are stable first order and shown in Table 1

Source: From the researcher by using Eviews program
The model was estimated using the ARDL method and the conitegration and long run form (CALRF), one lag time for the dependent variable and one lag time for the independent variables with no constant and direction to obtain the results shown in table (4), what is important in this estimate is that the CALRF, which is exactly like Johansson's fault-correction model, is different from the value of the coint Eq (-1), which is 0.90, with very high significance, where λ1 check the two conditions necessary in that it is negative and sufficient in its significance and explains that 90% of the short-term errors can be corrected by the unity of time, which is a year in order to return to a long-term equilibrium.The results of table (6) show that the model does not suffer from the problem of autocorrelation series according to the LM test as its statistical value appeared at the level of 0.3538, which makes us accept the null hypothesis that there is no problem of serial autocorrelation.The model does not suffer from the problem of heteroskedasticity.It has a probability value of 0.2860 which is greater than 0.05, which makes us accept the null hypothesis that the random error limit varies in the estimated model.Since the value of Jarque-Bera is greater than 5% normal distibution assurance model

Figure 1. Test the normal distribution of residuals Structural stability test results for the estimated ARDL model
The step after estimating the model formula is to test the structural stability of the short and long term coefficients, its mean, the data used in this research are free of structural changes over time, to achieve this, two tests are used: cumulative sum of recursive residua, (CUSUM) and cumulative sum of square recursive residual, (CUSUMS).The structural stability of the estimated coefficients of the UECM form of the ARDL model is achieved if the CUSUM and CUSUMSQ statistics are within the critical limits at a significant level of 5%.Hence, these coefficients are unstable if the diagram of the above two tests (6). Figure (2 and 3) shows that the estimated coefficients of the ARDL model are structurally stable over the period under study, confirming stability between the study variables and consistency in the model in the short and long run.

RESULTS AND DISCUSSION
The results of table (4), show the long-term relationship between the total domestic agricultural production and the independent variables, showing that the change in all these independent variables has a significant effect on the agricultural GDP (variable dependent) except X1 (subsidies to the agricultural sector) among the explanatory variables, a combined integration with the agricultural GDP index is integrated in the sense that there is a long-term relationship between these explanatory and dependent variables (agricultural GDP) and that there is a causal relationship in the short and long term moving from the explanatory variables to the variable, that mean, the model is stable, which means that the probability of these variables being effective is high in the long term.As the results showed • There is a negative and insignificant effect on the variable of supporting the agricultural sector in the long term after it was positive and significant at 10% in the short term.This is in line with what is recommended by the IMF and the World Bank in their reform programs for developing countries.Services and prices of factors of production and the need to work in real prices, which must economically cover the cost of production and abandon the policy to support crop prices and remove subsidies on agricultural inputs such as chemical fertilizers, seeds and pesticides as they represent a heavy burden on agricultural gross domestic product.It is worth mentioning that the negative impact of long-term support to the agricultural sector is consistent with several studies such as ( 14) and ( 17), as it is theoretically explained that producers are working to reduce the use of inputs as a result of ensuring a share of the income coming from the subsidy.This negative impact may be mainly due to the low productivity of the factors of production in the long term because of support, and the product may change its behavior and starts in the search for investment in activities that are subsidized is considered relatively less productive.In addition, the results showed there is a positive and very significant effect of the coefficient of the variable agricultural loans as an increase of 1% in this coefficient will lead to an increase of agricultural GDP by 9.6%.This confirms the role and importance of agricultural loans in the long term to revitalize the Iraqi agricultural sector.If agricultural loans work if they are best exploited to increase the production of farmers in the various projects for which they borrowed these amounts, which will be reflected positively on the increase in agricultural GDP, and long-term results are consistent with what can be the use of the correct loans, That the use of loans in the short term was not effective, which confirms that the adoption of lending institutions to take decisive action to guide the use of loans in their real purposes, especially long-term loans.Also the results showed a positive and very significant effect of the quantities received from the wheat crop and the quantities received from the rice crop.A 1% increase in the X3 parameter will increase the agricultural output by 2.26% and the increase of 1% in the X4 parameter will increase the agricultural output by 0.23%.The positive impact of these two variables clearly indicates the success of the policy of the government and the right of its actions in motivating farmers towards the delivery of quantities produced to warehouses and silos of the government as well as increased awareness of the producers of the need to take these procedures, which will positively reflect the level of self-sufficiency of the main crops to acceptable degrees and encouraging , which has been observed in recent years as the low food gap for major crops, especially wheat, has been shown to be in favor of higher agricultural GDP.Finally, the results showed a positive and significant effect of the parameter of the variable coefficient of net nominal protection of the wheat crop.The negative and significance relation between the variable of the net nominal protection coefficient for the rice crop and the agricultural GDP is expected because this crop is controlled by factors other than supporting the producers of this crop.The study recommends reducing subsidies and leaving price incentives operating within the market mechanism, reflecting the ability of the market to encourage agricultural production and to intervene in the pricing of agricultural products in a relative manner.The rice crop is linked to the development of successful water resources policies and programs, given that this crop is governed by the water component, which is a temporal and geographical component.Therefore, the only way to know what might happen is to predict through mathematical equations and digital models that document past events according to various scenarios in order to develop appropriate water policies are integrated with price policies formulated by the government for this strategic crop ‫كان‬ ‫اء‬ ‫سو‬ ‫قطاع‬ ‫كل‬ ‫اسيام‬ ‫مدى‬ ‫توضح‬ ‫االجمالي‬ ‫المحمي‬ ‫الناتج‬ ‫حسابات‬ ‫ان‬ ‫و‬ ‫الوطني،‬ ‫االقتصاد‬ ‫تكوين‬ ‫في‬ ‫نتاجيا‬ ‫اعي‬ ‫الزر‬ ‫القطاع‬ ‫تكوين‬ ‫في‬ ‫الصدارة‬ ‫مقام‬ ‫أ‬ ‫يتبو‬ ‫لم‬ ‫انو‬ ‫رغم‬ ‫الميمة‬ ‫القطاعات‬ ‫من‬ ‫في‬ ‫سيما‬ ‫ال‬ ‫النفطي‬ ‫القطاع‬ ‫مساىمة‬ ‫ضخامة‬ ‫بسبب‬ ‫االجمالي‬ ‫المحمي‬ ‫الناتج‬ ‫ا‬ ‫عودة‬ ‫شيدت‬ ‫التي‬ ‫االخيرة‬ ‫ات‬ ‫السنو‬ ‫الدولية‬ ‫النفطية‬ ‫السوق‬ ‫الى‬ ‫اق‬ ‫لعر‬ ‫في‬ ‫اعية‬ ‫الزر‬ ‫االقتصادية‬ ‫ات‬ ‫المتغير‬ ‫بعض‬ ‫اثر‬ ‫قياس‬ ‫البحث‬ ‫استيدف‬ .‫ىذه‬‫تمعبو‬ ‫ان‬ ‫يمكن‬ ‫الذي‬ ‫الدور‬ ‫وتحميل‬ ‫اعي‬ ‫الزر‬ ‫المحمي‬ ‫الناتج‬ ‫اجمالي‬ ‫النمو‬ ‫تعزيز‬ ‫خاللو‬ ‫من‬ ‫يمكن‬ ‫التي‬ ‫و‬ ‫الناتج‬ ‫ىذا‬ ‫نمو‬ ‫عمى‬ ‫ات‬ ‫المتغير‬ ‫اسة‬ ‫لدر‬ ‫الحديثة‬ ‫األساليب‬ ‫احد‬ ‫اتباع‬ ‫طريق‬ ‫عن‬ ‫نتائجو‬ ‫الى‬ ‫الوصول‬ ‫في‬ ‫الكمي‬ ‫األسموب‬ ‫عمى‬ ‫البحث‬ ‫.اعتمد‬ ‫اقي‬ ‫العر‬ ‫اعي‬ ‫الزر‬ ‫القطاع‬ ‫في‬ ‫ونموذج‬ ‫ات‬ ‫المتغير‬ ‫متعدد‬ ‫المشترك‬ ‫التكامل‬ ‫أسموب‬ ‫وىو‬ ‫السببية‬ ‫العالقة‬ ARDL ‫اخت‬ ‫و‬ ‫ات‬ ‫المتغير‬ ‫بين‬ ‫العالقة‬ ‫اتجاه‬ ‫لتحديد‬ ‫السببية‬ ‫العالقة‬ ‫بار‬ ‫البنك‬ ‫بيانات‬ ‫وعمى‬ ‫االقتصادية‬ ‫النظرية‬ ‫اضات‬ ‫افتر‬ ‫عمى‬ ‫معتمدا‬ ‫اسة‬ ‫الدر‬ ‫محل‬ ‫االقتصادية‬ ‫لإلحصاء‬ ‫باالستناد‬ ‫اعي‬ ‫الزر‬ ‫الناتج‬ ‫اجمالي‬ ‫في‬ ‫المؤثرة‬ ‫ات‬ ‫المتغير‬ ‫بعض‬ ‫تحديد‬ ‫تم‬‫وقد‬ ‫النات‬ ‫مؤشر‬ ‫من‬ ‫كل‬ ‫بين‬ ‫االجل‬ ‫طويل‬ ‫ا‬ ‫اثر‬ ‫ىناك‬ ‫ان‬ ‫الى‬ ‫البحث‬ ‫توصل‬ ‫وقد‬ ‫الموضوع‬ ‫ىذا‬ ‫تناولت‬ ‫التي‬ ‫ات‬ ‫المتغير‬ ‫وباقي‬ ‫اعي‬ ‫الزر‬ ‫المحمي‬ ‫ج‬ .‫بينيما‬ ‫االجل‬ ‫وقصيرة‬ ‫االجل‬ ‫طويمة‬ ‫سببية‬ ‫عالقة‬ ‫ىناك‬ ‫ان‬ ‫و‬ ‫اسة‬ ‫الدر‬ ‫قيد‬ ‫االقتصادية‬ ‫مفتاحية:‬ ‫كممات‬ ‫اعية‬ ‫الزر‬ ‫المضافة‬ ‫القيمة‬ ‫صافي‬ ‫المتابعة‬ ‫اقي‬ ‫لمبو‬ ‫اكمي‬ ‫التر‬ ‫المجموع‬ ، CUSUM ، ‫اعي‬ ‫الزر‬ ‫الثابت‬ ‫المال‬ ‫اس‬ ‫ر‬ ‫اكم‬ ‫تر‬ ‫اطروحة‬ ‫من‬ ‫مستل‬ ‫*البحث‬ ‫االول‬ ‫لمباحث‬ ‫اه‬ ‫الدكتور‬ *Received:13/1/2018, Accepted:15/4/2018

1 X: 2 X 3 X 4 X 6 X
Amounts of subsidies provided to the agricultural sector (in million Iraqi Dinars) : Agricultural loans (in million Iraqi dinars) : Amounts received from wheat crop (tones) : Amounts received from rice crop (tons) 5 X : Net Nominal Protection coefficient for Wheat Crop : Net Nominal Protection coefficient for Rice Crop == ti U : Random error term.s b , : Model parameters.The general formula of the ARDL model based on the UECM model and the BOND test proposed by (almusabah) 1 is composed of a dependent variable and K of the independent variables is(12):

Table 1 . Check the stability of time series UNIT ROOT TEST RESULTS TABLE (ADF) Null Hypothesis: the variable has a unit root
.