DR. NANO PRAWOTO. AGUS TRI BASUKI, SE., M.Si. (Lecturer Faculty of Economics and Business University Muhammadiyah Yogyakarta) ABSTRAKS

Transcription

1 MODEL ANALYSIS OF COMPOSITION OF PUBLIC EXPENDITURE ON ECONOMIC GROWTH IN WEB-BASED SUPPORT GOOD GOVERNANCE: EMPIRICAL STUDY CITY DISTRICT IN INDONESIA YEAR DR. NANO PRAWOTO (Lecturer Faculty of Economics and Business University Muhammadiyah Yogyakarta) AGUS TRI BASUKI, SE., M.Si (Lecturer Faculty of Economics and Business University Muhammadiyah Yogyakarta) ABSTRAKS This study aimed to analyze the quality of local government spending on economic growth. The method used in this study is the linear regression panel data. Spending area for health have an impact on economic growth, so that the local government must continue its program of Healthy Indonesia which is to protect the health of its citizens through BPJS or other health insurance. Education spending has no positive effect on economic growth in a region. It is caused by an increase in the education budget has not been matched by an increase in equity of education in certain areas, such as the lack of teachers, school buildings are not feasible and the lack of teaching and learning facilities. Kids are the future investment of the nation, so the government needs to increase the resources of children through free education at least 15 years old (elementary through high school), and for remote areas in 2020 should have a minimum of 12 years education. While the relationship between BPK's opinion and economic growth have an inverse relationship, meaning that if the BPK's opinion on the financial statements of local government improved had lower economic growth, this is due to the CPC simply checking the audit of financial statements is in accordance with the prescribed rules but did not see the performance but only saw whether the money spent already by the rules. The government through the minister of the interior had to seek enforcement of Performance-Based Budgeting, meaning that any public money spent should be coupled with a measurable performance that can push people's welfare. Keywords: economic growth, performance-based budgeting, and sosial welfare P a g e 1

2 A. BACKGROUND Basically regional development can not be released to national development, one of Indonesia's national development goals is to create economic growth and equitable distribution of the fruits of development, including the equitable distribution of income among regions (regions). To achieve the above objectives is not easy work because the general economic development of a region closely linked to the economic potential and its regional characteristics. To reach long-term goals and objectives essential to the building, the Indonesian national development the next five years need to prioritize in order to achieve food sovereignty, energy sufficiency and management of maritime and marine resources. Along with that, the construction of the next five years should also be increasingly lead to the condition of improving the welfare of a sustainable, citizens personality and spirit of mutual cooperation, and people have a harmony between social groups, and the posture of the economy increasingly reflect quality growth, which is inclusive, broad-based, based on the benefits of human resources and the ability of science and technology while moving toward the balance between economic sectors and between regions, and increasingly reflects the harmony between man and the environment (RPJMN ). In order to achieve national goals, the Indonesian people are faced with three main issues, namely: (1) the decline of state authority; (2) the weakening of the joints of the national economy; and (3) the spread of intolerance and national identity crisis Nations Economic Weakness joints. Weak joints nation's economy looks of unresolved problems of poverty, social inequality, disparities between regions, environmental damage due to the exploitation of natural resources is excessive, and dependence in terms of food, energy, finance, and technology. Countries unable to capitalize on the wealth of nature is very large, well embodied (tangible) and non-physical (intangible), for the welfare of its people. Hopes for strengthening the joints of the nation's economy becomes increasingly distant when the state has no power to guarantee the health and quality of living for its citizens, failed to reduce inequality and inequality in national income, perpetuates dependence on foreign debt and the provision of food that rely on imports, and not responsive in addressing the issue of energy crisis due to the domination of the means of production and global corporate capital and reduced national oil reserves. Regional economic development has two key objectives increasing the number and types of employment opportunities for local communities, in an effort to achieve that goal, local governments and communities should jointly take the initiative to build the region. Therefore, local governments must seek to use existing resources in the area with as appropriate for the prosperity of the masses and drive the economy forward. Fiscal policy is an economic policy in order to drive the economic conditions to be better with the change of government revenue and expenditure. With planning and supervision countries / regions will have an impact on economic conditions is expected that the welfare of society. To reduce the leakage of the state budget or the budget, the government needs to monitor with the help of the KPK and BPK. In addition to the planning and supervision of P a g e 2

3 APBN / APBD in influencing economic conditions, investment can also be relied upon to create a development strategy Pro Growth, Pro-Poor and Pro Job. Arsyad explained that any attempt at regional economic development has the main goal to increase the number and types of employment opportunities for local communities. In an effort to achieve that goal, local governments and communities should jointly take the initiative to build the region. Local government and community participation and the use of existing resources attempt to inventory the potential of existing resources to design and build the local economy. Seriousness of the government in establishing this area is measured by the existence of a system of government known as regional autonomy. To support this government passed Law No. 22 of 1999 on Regional Government which was later revised to Act 32 of 2004 and Law No. 25 of 1999 on the financial balance between central and local governments are then revised by the Act No. 33 of The law is the foundation for the region to develop their regions independently to rely more on the capabilities and potential of the area. This law also gives greater authority (local discretion) to the region to design development programs in accordance with the wishes of the local community (local needs). This study is expected to prove the role of local government spending, especially in the fields of education, health, and Fisheries and Maritime Affairs, Investments and BPK opinion against LKPD in stimulating economic growth, so as to create effectiveness in regional economic development and the creation of good governance. B. PROBLEM FORMULATION From the background of the problems above, we can formulate the problem as follows: 1. How much influence between regional revenue to local economic growth. 2. How much influence between regional revenue for education to regional economic growth. 3. How much influence between government expenditures for education to regional economic growth. 4. How much influence between Government Spending For Health to regional economic growth. 5. How much influence among the population of the region's economic growth. 6. How much influence between Audit Board against LPKD Opinion on economic growth area. C. RESEARCH OBJECTIVES The purpose of this study is: 1. Want to know the relationship between regional revenue to local economic growth. 2. Want to know the relationship between the government expenditures for education to regional economic growth. 3. Want to know the relationship between the Government Spending For Health to regional economic growth. P a g e 3

4 4. Want to know the relationship between the number of population on economic growth in the region. 5. Want to know the relationship between Audit Board against LPKD Opinion on economic growth area. D. URGENCY OF RESEARCH Routine budget play an important role to support the mechanisms of the system of government as well as efforts to increase efficiency and productivity, which in turn will support the achievement of the goals and objectives of each stage of development. Meanwhile, development expenditure is intended to finance the development programs whose budgets are always adjusted to the funds that have been mobilized. Government spending in real terms can be used as an indicator of the magnitude of government spending financed by the government and how a proportion of national income. The bigger and more activities, the greater the government's public expenditure is concerned. But we should realize that the proportion of government expenditure to gross national income (GNP) is a very rough measure of the role of Government in an economic activity. As an illustration, the fiscal policy implemented by the government are often virtual in the short term or not felt by the public because of the economic activity in the short term is relatively insignificant, and in the long term, the dimensions of social justice economies of poor arrangement fiscal policy will obviously burden the people from various walks of life, The problem is that the details of the policy that is in APBN/APBD often do not show the direction of policy and become guideliness program to provide opportunities for the stimulation of economic activity and the private sector. The study is expected to be useful in assessing the effectiveness of local government spending, especially in encouraging regional economic growth, so as to create a development strategy that is pro-poor, pro-jobs and pro-growth. E. THEORETICAL 1. Understanding Economic Growth According to Prof. Simon Kuznets, define economic growth as a long-term increase in a country's ability to provide more and more kinds of economic goods to its citizens. This ability to grow in accordance with advances in technology, and adjustment of institutional and ideological needs. This definition has three (3) components: First, the economic growth of a nation to improved continuous inventories; second, advanced technology is a factor in the economic growth that determines the degree of growth ability in supplying various kinds of goods to the population; Third, the use of technology is widely and efficiently requires adjustments in the institutional field and the ideology that the innovation generated by human knowledge can be utilized appropriately (Jhingan, 2000: 57). Economic growth is a process of increase in output per capita in the long term, where the emphasis on three things: process, output per capita and long term. Economic growth is a "process" is not an economic picture at a time. Here we see the dynamic aspect P a g e 4

5 of an economy, which is to see how an economy grow or change over time. The emphasis on changes or development itself. a. The views Boediono Economic growth is also related to the increase "output per capita". In this sense the theory must include theories about the growth of GDP and theories on population growth. Because only when these two aspects are explained, the per capita output growth can be explained. Then the third aspect is economic growth in the long-term perspective, that if during the period of time long enough that output per capita shows an upward trend (Boediono, 1992: 1-2). b. The views Adolp Wagner According to empirical observations Adolp Wagner against European countries, the United States and Japan in the 19th century show that government activity in the economy tends to increase (law of ever increasing state activity). Wagner measure of the ratio of government expenditure to revenue nasional. According to the understanding Wagner, there are some things that cause government spending always increases, namely, the increasing demands of protection of security and defense, rising income levels, urbanization accompanying economic growth, the development of democracy and inefficient bureaucracy that accompanies government development. According to Wagner legal, economic growth will cause the relationship between industries, industrial-society, and so will be more complicated and complex that the potential market failure occurs and the greater negative externalities. Correspondingly, as shown in figure II.1. relative role of government will increase (Mangkoesoebroto, 1993: 171). c. The views W.W. Rostow and Musgrave W.W. Rostow and Musgrave linking government spending to the stages of economic development. In the early stages of economic development, government investment ratio to total invetasi, or in other words the ratio of government expenditure to national income is relatively large. This is because at this early stage the government should provide the infrastructure. At the intermediate stage of economic development, government investment is still needed to spur economic growth in order to remain able to take off. Along with that portion of the private sector is also to be increased. The government's role is still mostly caused by a lot happens at this stage of market failure that caused by economic development itself. Many cases of negative externalities, such as environmental pollutionare demanding the government to intervene to overcome them. In a development process according to Musgrave, the ratio of total investment to national income grew, but the ratio of public investment to national income will shrink. Meanwhile Rostow found at an advanced stage of development, the transitional government activity occurred, and the provision of economic infrastructure kepengeluaran-spending for P a g e 5

6 social services such as health and education. Rostow and Musgrave, like Wagner, underlying earnings also based on the observations of the experience of economic development in many countries. d. The views Keynes Balance of national income identity Y = C + I + G + X - M is a source of legitimacy of the Keynesian view of the relevance of government intervention in the economy (Dumairy 1996: 161). Many considerations underlying the decision to regulate spending. The government is not enough to achieve the ultimate goal of any policy spending, but should also take into account the intermediate targets to be enjoyed or affected by the policy. Enlarge spending with the aim solely to increase national income or job opportunities is not sufficient, but must also be taken into account who gets hired or increase revenue. The government also needs to avoid in order to increase its role in the economy does not weaken private sector activity. Public economic experts have long paid attention to the investigation of the relationship between government spending with economic growth rates since they realize that government spending plays a very important in the economy of a country either at low or high income countries. F. LITERATURE Paolo Mauro, 1995, Corruption and Growth, hypothesis used in the study was the alleged corruption has a negative correlation with investment and economic growth. The findings could Corruption reduces investment, thus lowering economic growth. N. Gregory Mankiw, David Romer, David N. Weil, 1992, the data used are annual data from 1960 to 1985 period. Data collected include real incomes, government and private consumption, investment, and a population of almost all the world besides the centrally planned economy. They measure n as the average growth of the working age population, which is defined working age population aged 15 to 64. Their work is of poor countries tend to grow faster than rich countries. Suleiman U.S. Aruwa, 2012, public finances and economic growth in Nigeria, this research using annual time series data of the Federal Government of Nigeria, the results of research there is a long-term relationship between government spending and the GDP, and public expenditure with revenue for the case of Nigeria. Kwabena Gyimah-Brempong, 2002, Corruption, economic growth, and income inequality in Africa, In this paper defines corruption as the use of public office for private gain. Results of research corruption reduces the rate of revenue growth. The increase in the corruption index unit reduces GDP growth rate of between 0.75 and 0.9 percentage points, and the per capita income of between 0.39 and Corruption lower levels of per capita income growth directly by lowering the productivity of existing resources and indirectly through reduced investment. P a g e 6

7 Kevin Sylwester, 2000, Income inequality, education expenditures, and growth, The aim of this study was to determine whether income inequality can reduce growth by increasing public spending on education. Results of research does help explain why increased levels of education do not have a positive influence, and direct influence on economic growth. Although the benefits arising from the increase in human resources are higher, there are also costs associated with the increase in human resources and these costs can provide a relationship why income inequality lowers economic growth. G. Framework The framework is used as a guide or as a picture of this line of thought in focus on the research objectives. Economic growth can be affected by several factors. If we look at the framework below, it can be seen that economic growth can be affected by several factors, among others, the original income (PAD), expenditure on health (KESH), expenditures for Education (EDUC), Total population (PENDDK) and Opinion BPK against Local Government Financial Reporting. the original income (PAD) expenditure on health (KESH) expenditures for Education (EDUC) Economy Growth Total population (JP) Opinion BPK against Local Government Financial Reporting P a g e 7

8 H. Hypotheses Research From a theoretical framework and previous research then can we derived hypotheses as follows: 1. Allegedly there is a positive relationship between regional revenue for education to regional economic growth. 2. Presumably there is a positive correlation between government spending for education to regional economic growth. 3. Allegedly there is a positive correlation between government spending for health on economic growth area. 4. Presumably there is a positive correlation between the number of population on economic growth in the region. 5. Presumably there is a negative relationship between the CPC against LPKD Opinion on economic growth area. I. Analysis Tools Regression was used in this research is regression with panel data. Panel data is a combination of time series and cross section. According to Agus Widarjono (2009) use panel data in an observation has several benefits. First, the panel data which is a combination of two data time series and cross section able to provide more data so that it will produce a degree of freedom is greater. Second, combining information from data time series and cross section can solve problems that arise when there is a problem disappearances variables (omitted-variable). Hsiao (1986), noted that the use of panel data in economic research has several major advantages compared to the data type of cross section and time series. First, it can give researchers a large number of observations, increasing the degree of freedom (degrees of freedom), the data has a great variability and reduce collinearity between explanatory variables, which can produce an efficient econometric estimation. Second, the panel data can provide more information that can not be given only by the cross section or time series only. And Third, the panel data can provide better solution than the dynamic changes in the inference cross section data. According Wibisono (2005) benefit panel data regression: First. Panel data is able to take into account the heterogeneity of individuals explicitly turn by allowing individual specific variables. Second. The ability to control this heterogeneity makes the subsequent panel data can be used to test and build more complex behavioral models. Third, panel data were based on the observation cross section repeatedly (time series), so that the panel data method suitable for use as a study of dynamic adjustment. Fourth, the high number of observations have implications on the data that is more informative, more varied, and collinearity between the data on the wane, and the degrees of freedom (degree of freedom/df) higher so the estimation results can be obtained more efficiently. Fifth, the panel data can be used to study models of complex behavior. And Sixth, panel data can be used to minimize the bias that may be caused by aggregation of individual data. P a g e 8

9 J. Panel Data Regression Model Panel Regression Model of the above title as follows: Y = α + b1x1 it + +b2x2 it + e Where : Y = dependent variable α = Constant X1 = The independent variable 1 X2 = The independent variable 2 b (1... 2) = regression coefficient of each independent variable e = Error term t = Time i = Company K. Regression Model Estimation Method Panel In the regression model estimation method using panel data can be performed through three approaches, among others: 1. Common Effect Model Panel data model approach is the simplest because it combines data time series and cross section. In this model does not pay attention to the dimension of time as well as individuals, so it is assumed that the behavior of the data the same company in different periods. This method can use the approach Ordinary Least Square (OLS) or a least squares technique to estimate the panel data model. 2. Fixed Effect Model This model assumes that differences between individuals can be accommodated on the difference intercept. To estimate the Fixed Effects panel data models using the technique of dummy variables to capture the difference between the company's intercept, the intercept differences may occur because of differences in the work culture, managerial, and incentives. However slop is equally between the companies. The estimated model is often called the technique of Least Squares Dummy Variable (lsdv). 3. Random Effect Model This model will be estimating panel data where disturbance variables may be interconnected across time and between individuals. Random Effect on the model intercept differences are accommodated by the error terms of each company. Advantages of using the Random Effect models that eliminate heteroscedasticity. This model is also called the Error Component Model (ECM) or technique Generalized Least Square (GLS). P a g e 9

10 Selection Model To select the most appropriate model used in managing the data panel, there are several tests that can be made are: 1. Test Chow Chow test that is testing to determine the model Fixed Effect Effet or Random most appropriately used in estimating panel data. 2. Test Hausman Hausman test is a statistical test to choose whether or models Fixed Effect Random Effect is most appropriately used. 3. Test Lagrange Multiplier To find out if the model is better than the Random Effect or Effect Common method (OLS) test was used Lagrange Multiplier (LM). L. Framework Research Obyek Dependent Var. Economy Growth Independent var : 1. the original income (PAD), 2. expenditure on health (KESH), 3. expenditures for Education (EDUC), 4. Total population (PENDDK) and 5. Opinion BPK against Local Government Financial Reporting. Estimasi Panel Data Model Common Effect Fixed Effect Random Effect Selected Panel Data Model LM test Chow test Hausman test Klasical asumption Source : Gujarati, 2003 Intrepretation Figure 1: The framework P a g e 10

11 M. DATA ANALYSIS 1. GRDP Definition of domestic / regional herein can be a Provincial or District / City. Economic transactions that will be calculated are transactions that occurred in the domestic territory of a region regardless of whether the transaction is done by the community (resident) from the region or other community (non-resident). No Table 1: GRDP in 31 districts / cities in Sumatra and Java GDRP Commentary Average Ecomy Growth 1 Kab. Aceh Besar 2,618 2,736 2,860 2, Kab. Aceh Tenggara Kab. Bireuen 2,775 2,930 3,067 3, Kab. Asahan 5,680 6,252 6,532 7, Kab. Dairi 2,159 2,276 2,400 2, Kab. Deli Serdang 15,389 16,322 18,410 19, Kab. Agam 3,280 3,503 3,726 3, Kab. Pasaman 1,453 1,542 1,637 1, Kota Padang 12,792 13,638 14,517 15, Kota Payakumbuh ,061 1, Kota Sawahlunto Kab. Bengkalis 30,398 29,894 28,038 27, Kab. Kampar 9,730 10,247 10,723 11, Kab. Merangin 1,267 1,349 1,436 1, Kab. Tebo 971 1,037 1,104 1, Kab. Lahat 2,892 3,068 3,237 3, Kab. Musi Rawas 3,859 4,063 3,155 2, Kab. Muara Enim 8,940 9,393 8,261 8, Kab. Ogan Komering Ilir 3,544 3,777 4,021 4, Kab. Ogan Komering Ulu 3,012 3,232 3,435 3, Kab. Bengkulu Selatan Kab. Bengkulu Utara Kab. Kaur Kab. Kepahiang Kab. Lampung Tengah 6,587 7,007 7,436 7, Kab. Sukabumi 2,042 2,150 2,255 2, Kab. Brebes 5,781 6,082 6,390 6, Kab. Magelang 4,292 4,543 4,797 5, Kota Surakarta 5,412 5,743 6,081 6, Kab. Kulon Progo 1,869 1,963 2,062 2, Kab. Blitar 6,082 6,468 6,868 7, Source: BPS (various issues) P a g e 11

12 2. The development of PAD Local revenue (PAD) is all the revenue that the area from sources within its own territory levied by local regulations in accordance with the legislation in force (Halim, 2004: 96). Regional income sector plays a very important, because through this sector can be seen the extent to which the region can finance the activities of the government and regional development. Increased revenue (PAD) absolutely must be done by the local government to be able to finance their own needs, so that the dependence of local government to the central government on the wane and eventually the area can be independent. In Act No. 33 of 2004 on the financial balance between the central government and local governments in Chapter V (five) number 1 (one) mentioned that the local revenue derived from: a. Local tax According to Law No. 28 of 2009 Local Taxes, hereinafter called Tax, is a mandatory contribution to a region that is owed by private persons or entities that are enforceable under the Act, by not getting the rewards directly and used for the purposes of Regions for the overall prosperity people. Based on Law No. 28 of 2009 tax district/city is divided into several following, hotel tax, restaurant tax, amusement tax, advertisement tax, street lighting tax, Tax Mineral instead of metals and rocks, parking tax, Ground Water Tax, Tax Swallow's Nest, Land and Building Tax in Rural and Urban Areas and Customs Tax on Acquisition of Land and Buildings. b. Levies According to Law No. 28 of 2009 as a whole there are 30 types of user charges can be levied by the regions which are grouped into three categories levy, namely general service levies, charges for services, and certain licensing levies. c. Results-owned wealth management separated areas Results-owned wealth management area that separated the reception area from the separated areas of wealth management. Law number 33 of 2004 classifies the type of results the wealth management separated areas, specified according to according to the object of income that covers part of the return on equity investment in companies owned by regional/local enterprises, part of the return on its investment in state-owned enterprises/state and part of the return on equity investment in privately-owned companies and community groups. d. Other Local Revenue legitimate Law No. 33 of 2004 explains the regional revenue legitimate, provided for budgeted reception area is not included in this type of tax and wealth management outcomes separated areas. P a g e 12

13 Table 2: Local Revenue in 31 districts / cities in Sumatra and Java No Commentary Local Revenue Growth Average 1 Kab. Aceh Besar 49,580 54,798 76,107 65, Kab. Aceh Tenggara 11,910 25,324 25,621 45, Kab. Bireuen 16,817 22,942 93, , Kab. Asahan 31,844 37,895 53,692 59, Kab. Dairi 17,673 20,912 29,933 24, Kab. Deli Serdang 213, , , , Kab. Agam 37,894 41,573 49,954 60, Kab. Pasaman 25,782 33,037 46,024 40, Kota Padang 149, , , , Kota Payakumbuh 44,561 50,709 54,178 60, Kota Sawahlunto 36,382 34,888 37,105 43, Kab. Bengkalis 206, , , , Kab. Kampar 96, , , , Kab. Merangin 39,648 30,113 44,396 47, Kab. Tebo 19,538 26,939 34,331 35, Kab. Lahat 70,066 70,938 78,313 79, Kab. Musi Rawas 65,428 73,018 75,367 81, Kab. Muara Enim 86, , , , Kab. Ogan Komering Ilir 47,387 54,618 68,701 62, Kab. Ogan Komering Ulu 42,120 41,429 44,680 68, Kab. Bengkulu Selatan 14,758 18,911 25,454 19, Kab. Bengkulu Utara 19,099 24,694 30,114 37, Kab. Kaur 6,811 7,782 10,339 12, Kab. Kepahiang 11,990 13,896 19,468 17, Kab. Lampung Tengah 50, ,060 81, , Kab. Sukabumi 151, , , , Kab. Brebes 78, , , , Kab. Magelang 90, , , , Kota Surakarta 181, , , , Kab. Kulon Progo 53,752 74,029 95,992 92, Kab. Blitar 76,191 95, , , Source: Ministry of Finance 3. The development of Government Health Spending Government spending is one aspect of the use of economic resources which are directly controlled by the government and indirectly owned by the public through tax payments. In general, government spending will increase in line with the increase in economic activity of a country. This situation can be explained in the rule known as P a g e 13

14 No Wagner's Law, which is the positive correlation between government spending to the level of national income. However, a large increase in government spending is not necessarily a good result against economic activity. Table 3: Government Health Spending in 31 districts / cities in Sumatra and Java Commentary Government Health Spending Growth Average 1 Kab. Aceh Besar 76,645 88, , , Kab. Aceh Tenggara 49,352 61,578 80, , Kab. Bireuen 68,262 87, , , Kab. Asahan 71,845 73,872 88,562 96, Kab. Dairi 53,226 66, ,272 87, Kab. Deli Serdang 157, , , , Kab. Agam 85,700 77,002 89, , Kab. Pasaman 19,767 63,481 86, , Kota Padang 51,494 94, , , Kota Payakumbuh 56,558 61,768 69,071 77, Kota Sawahlunto 29,924 48,750 69,323 80, Kab. Bengkalis 77, , , , Kab. Kampar 196, , , , Kab. Merangin 41,073 61,779 61,825 89, Kab. Tebo 156,129 45,981 60,355 64, Kab. Lahat 78, , , , Kab. Musi Rawas 41, , , , Kab. Muara Enim 46, , , , Kab. Ogan Komering Ilir 88, , , , Kab. Ogan Komering Ulu 68,943 80, , , Kab. Bengkulu Selatan 45,585 69,174 74,456 78, Kab. Bengkulu Utara 23,196 66,836 96,263 96, Kab. Kaur 70,959 34,071 35,995 47, Kab. Kepahiang 85,484 37,983 51,699 58, Kab. Lampung Tengah 88,101 11, , , Kab. Sukabumi 59, , , , Kab. Brebes 110, , , , Kab. Magelang 113, , , , Kota Surakarta 90,301 98, , , Kab. Kulon Progo 85, , , , Kab. Blitar 154,304 78,795 86,255 89, Source: Ministry of Finance P a g e 14

15 4. The development of the Education Sector Government Spending Education has an important role in the process of development of a region. Advancement of education will have an impact on improving labor productivity, increase the productivity impact on cost efficiency of production, and this will encourage the economic growth of a region. No Table 4: Education Sector Government Spending in 31 districts/cities in Sumatra and Java Education Sector Government Spending Commentary Growth Average 1 Kab. Aceh Besar 276, , , , Kab. Aceh Tenggara 176, , , , Kab. Bireuen 350, , , , Kab. Asahan 405, , , , Kab. Dairi 256, , , , Kab. Deli Serdang 789, ,546 1,053,278 1,225, Kab. Agam 207, , , , Kab. Pasaman 147, , , , Kota Padang 251, , , , Kota Payakumbuh 181, , , , Kota Sawahlunto 117, , , , Kab. Bengkalis 226, , , , Kab. Kampar 173, , , , Kab. Merangin 226, , , , Kab. Tebo 482, , , , Kab. Lahat 241, , , , Kab. Musi Rawas 115, , , , Kab. Muara Enim 190, , , , Kab. Ogan Komering Ilir 369, , , , Kab. Ogan Komering Ulu 368, , , , Kab. Bengkulu Selatan 123, , , , Kab. Bengkulu Utara 96, , , , Kab. Kaur 262, , , , Kab. Kepahiang 547, , , , Kab. Lampung Tengah 555, , , , Kab. Sukabumi 390, , , , Kab. Brebes 686, , , , Kab. Magelang 550, , , , Kota Surakarta 402, , , , Kab. Kulon Progo 618, , , , Kab. Blitar 714, , , , Source: Ministry of Finance P a g e 15

16 5. Developments Population The development of the population of an area will impact on the activity of economic activity in a region. If the increase in population coupled with an increase in reliable human resource then it will have an impact on increasing the efficiency of production costs, but on the contrary if population growth is not accompanied by an increase in human resources, it will cause social problems, namely an increase in unemployment and an increase in unemployment have an impact on the increase in poverty and poverty will decrease economic growth. No Commentary Table 5: Total Population in 31 districts/cities in Sumatra and Java Total Population Growth Average 1 Kab. Aceh Besar 359, , , , Kab. Aceh Tenggara 182, , , , Kab. Bireuen 397, , , , Kab. Asahan 677, , , , Kab. Dairi 273, , , , Kab. Deli Serdang 1,816,205 1,841,450 1,867,046 1,892, Kab. Agam 461, , , , Kab. Pasaman 256, , , , Kota Padang 845, , , , Kota Payakumbuh 118, , , , Kota Sawahlunto 57,654 58,427 59,210 60, Kab. Bengkalis 512, , , , Kab. Kampar 707, , , , Kab. Merangin 339, , , , Kab. Tebo 303, , , , Kab. Lahat 375, , , , Kab. Musi Rawas 533, , , , Kab. Muara Enim 727, , , , Kab. Ogan Komering Ilir 738, , , , Kab. Ogan Komering Ulu 329, , , , Kab. Bengkulu Selatan 145, , , , Kab. Bengkulu Utara 262, , , , Kab. Kaur 109, , , , Kab. Kepahiang 127, , , , Kab. Lampung Tengah 1,185,943 1,200,886 1,216,018 1,231, Kab. Sukabumi 303, , , , Kab. Brebes 1,748,383 1,762,719 1,777,174 1,791, Kab. Magelang 1,191,615 1,201,386 1,211,237 1,221, Kota Surakarta 503, , , , P a g e 16

17 No Commentary Total Population Growth Average 30 Kab. Kulon Progo 393, , , , Kab. Blitar 132, , , , Source: BPS (various issues) 6. The development of BPK opinion of the local Government Financial Statements According to the mandate of Article 18 of Law Number 15 Year 2004 concerning Management and Accountability of State Finance. According to these provisions, the CPC shall deliver IHPS to representative institutions as well as the president, governors, regents and mayors not later than three months after the expiration of the term concerned. No Table 6: BPK opinion of the local Government Financial Statements in 31 districts/cities in Sumatra and Java Commentary BPK opinion of the local Government Financial Statements Kab. Aceh Besar WDP WTP WTP WTP 2 Kab. Aceh Tenggara WDP WDP WDP WDP 3 Kab. Bireuen WDP WDP WDP WTP 4 Kab. Asahan WDP WDP WDP WTP-DPP 5 Kab. Dairi WDP WDP WDP WTP 6 Kab. Deli Serdang TMP TW TMP WDP 7 Kab. Agam WDP WTP-DPP WTP-DPP WTP 8 Kab. Pasaman WDP WDP WTP-DPP WTP-DPP 9 Kota Padang WDP WTP-DPP WDP WTP-DPP 10 Kota Payakumbuh WDP WDP WDP WTP-DPP 11 Kota Sawahlunto WDP WDP WDP WDP 12 Kab. Bengkalis TMP WDP WTP-DPP WTP-DPP 13 Kab. Kampar WDP WDP WDP WDP 14 Kab. Merangin WDP WDP WDP WDP 15 Kab. Tebo WDP WDP WDP WDP 16 Kab. Lahat WDP WDP WDP WTP-DPP 17 Kab. Musi Rawas WDP WDP WDP WTP 18 Kab. Muara Enim WDP WDP WTP-DPP WTP 19 Kab. Ogan Komering Ilir WTP WTP WTP WTP 20 Kab. Ogan Komering Ulu WDP WDP WDP WDP 21 Kab. Bengkulu Selatan WDP WDP WDP WTP 22 Kab. Bengkulu Utara WDP WTP WTP WTP 23 Kab. Kaur WTP WTP WTP WTP P a g e 17

18 No Commentary BPK opinion of the local Government Financial Statements Kab. Kepahiang WDP TMP WDP WDP 25 Kab. Lampung Tengah WDP WTP WDP WDP 26 Kab. Sukabumi WDP WDP WDP WDP 27 Kab. Brebes WDP WDP WDP WDP 28 Kab. Magelang WDP WDP WDP WDP 29 Kota Surakarta WTP WTP WTP WTP 30 Kab. Kulon Progo WDP WDP WTP-DPP WTP-DPP 31 Kab. Blitar WDP WDP WDP WDP Source: Ministry of Finance N. Analysis and Discussion 1. Quality Data Test Here is the output test results using the Test Park heteroscedasticity shown in the table below Table 7 : Park Test of heteroscedasticity Variabel Prob. C LOG(PAD?) LOG(KESH?) LOG(EDUC?) LOG(PENDDK?) OPINI? Source: Data processed information: *** = signifikan 1% ** = signifikan 5% * = signifikan 10% From Table 4, it can be concluded that the data used as independent variables free from heteroscedasticity. Test Multicollinearity is a state where the independent variables in a multiple regression model found a correlation (correlation) between one another. Multicollinearity test aims to test whether the regression found the correlation. In the event of multikolinearitas, the regression coefficients of the independent variables will not be significant and have a high standard error. The smaller the correlation between independent variables, the regression model, the better (gujarati, 2003). Of attachments can be seen that the correlation coefficient between independent variables no bigger than [0.9] so that the data in this study there was no trouble multikolinearitas. P a g e 18

19 2. Analysis of Top Model In the analysis of panel data model, there are three kinds of approaches that can be used, namely least squares approximation (ordinary/pooled least square), fixed effects approach (fixed effect), and the approach of random effects (random effect). Statistical tests to select the first model is the Chow test to determine whether the method of least squares or Fixed Pooled effect that should be used to create a panel data regression. Selection of this model using the best analytical test more fully described in the following table. Table 8 : Result Regression Dependent Variable : Model (LOG(PDRB)) Common Effect Fixed Effect Random Effect Konstanta Standar error Probabilitas LOG(PAD) Standar error Probabilitas LOG(KESH) Standar error Probabilitas LOG(EDUC) Standar error Probabilitas LOG(PENDDK) Standar error Probabilitas Dummy Standar error Probabilitas R F statistik Probabilitas **** **** **** Source: Data processed Description: **** = significant 1% *** = significant 5% ** = significant 10% Based on the test model specifications that have been made from two analyzes performed by using Test Likelihood and Hausman Test both suggested to use a Fixed Effect, and from a comparison test best choise the regression model used in estimating Effect of revenue, expenditure on education, expenditure health, population, and KPK opinion on economic growth in the districts/cities in Sumatra and Java is the Fixed Effects Model. Fixed Effect Model was chosen because it has a probability of each independent variable on the Fixed Effect model is more significant than the Random Effect Model or Common Effect Model that each independent variable is not significant so much better model ie Fixed Effect Model. P a g e 19

20 Selection of panel data methods of testing performed on all sample data, Chow test is done to choose the method of panel data test between Pooled least square method or Fixed Effect. If the value of F statistic at significant Chow test, the Hausman test will be conducted to choose among methods Fixed Effect or Random Effect. Hausman test results with a probability value of less than Alpha is significant, meaning Fixed Effect method chosen to process the data panel. Selection method of testing is done by using the option Fixed Effects and Random Effect and combine both crosssection, period, or a combination of cross-section/period. 3. Data Testing Method Selection Panel a. Chow Test (Test likelihood) Chow test is a test to determine the best model among the Fixed Effect Model with Common / Pool Effect Model. If the results declared null hypothesis receives the best model to use is Common Effect Model. However, if the result is declared then reject the null hypothesis that the model used is the Fixed Effects Model, and testing will continue to Hausman test. Fn-1, nt, n-k = / = F-table = ; df (n-1, nt-n-k) = 5%; (31-1, ) = 5%; (30, 88) = The results of the calculation of the F-count obtained at , while the F-table of the numerator 30 and 88 on denumenator: 5% is From the above can be concluded that the hypothesis H0 is rejected because the F-count is greater than F- table ( > 1.649), so that the model used in this study is the Fixed Effects Model. Based Chow Test table above, these two probability values Cross Section F and Chi Square smaller than 0.05 Alpha so reject the null hypothesis. So according to the Chow test, the best model used is the model by using Fixed effect. Based on the results of the Chow test reject the null hypothesis, the test data is continuing to test Hausman. b. Hausman test Hausman test is a test to determine the method of use with Fixed Effect Random Effect. If the result of the Hausman test the null hypothesis that they received the best model to use is a model Random Effect. However, if the result is declared then reject the null hypothesis that the model used is a model Fixed Effect. P a g e 20

21 Tabel 9 : Hausman Test Test Summary Chi-Sq. Statistic Chi-Sq.d.f Prob. Cross-section random 69, Source : Data Processed According to the table Hausman test, the probability value is a random cross-section smaller than Alpha 0.05 to reject the null hypothesis. So according to Hausman test, the best model used is the model by using Fixed Effect. c. Panel Data Model Estimation Results Fixed Effect Model (FEM) Based on the test specifications of the model that has been done as well as from a comparison of the best value then the regression model used is the Fixed Effects Model. Fixed Effect Model (FEM) is a panel data estimation technique using dummy variables to determine their intercept differences between cross section. The following table shows the results of the estimation of data with the number of observations by 31 regencies / cities during the period (4 years). Table 10 : Fixed Effect Model Dependent Variable : (LOG(PDRB)) Model Fixed Effect Konstanta Standar error Probabilitas 0,000 LOG(PAD) Standar error Probabilitas LOG(KESH) Standar error Probabilitas LOG(EDUC) Standar error Probabilitas LOG(PENDDK) Standar error Probabilitas 0,000 Dummy Standar error Probabilitas R P a g e 21

22 F statistik Probabilitas 0,00 Durbin-Watson stat Source : data prosessed From the estimation above, it can be made model of panel data analysis of the factors affecting economic growth case study in Sumatra and Java interpreted as follows: LOG (PDRB) = β0 + β1*log(pad) + β2*log(kesh) + β3*log(educ) + β4*log(pend) + β5*dummy + et LOG (PDRB) = -26, ,055386*LOG(PAD)*** + 0,007851*LOG(KESH)* 0,00647*LOG(EDUC) + 2,62398*LOG(PEND)**** 0,00893*Dummy*** + et Where : **** Where significant at alpha 1 percent *** Where significant at alpha 5 percent ** Significant at alpha where 10 percent * Where significant at alpha 20 percent β0 = β1 = Β2 = Β3 = Β4 = means that when all the independent variables (PAD, education spending, health spending, population and opinion towards LKPD CPC) held constant or does not change there will be no economic growth (equal to antilog approaching enol) in Sumatra and Java can be interpreted any other factors considered to remain so if the original income increased by 1 percent will result in an increase in economic growth of percent can be interpreted any other factors considered to remain so if the Health Spending increased by 1 percent will result in an increase in economic growth of percent. Increased health costs have an impact on public health and the impact on productivity improvements population can be interpreted other factors held constant so if education spending increased by 1 percent will result in an increase in economic growth of percent. This means that education has an important role in stimulating local economic growth, this increase in education impact on improving human resources and ultimately increase the productivity of labor. But when seen from the results of the partial regression seen that education spending has no effect on economic growth this is due to the low numbers of the old school in Indonesia, the average length of less than 9 years of school or have not graduated from high school can be interpreted any other factors considered to remain so if P a g e 22

23 Β5 = the population increased by 1 percent will result in increased economic growth of 2.62 percent. In the short term the growth in population has benefits in boosting economic growth, but in the long term population growth can hamper economic growth of a region can be interpreted other factors held constant so if there is an inverse relationship between the CPC opinion with economic growth, its mean if the BPK's opinion on the financial statements of local governments improved it lowers down economic growth, this is due to the CPC only check the financial statement audit whether in accordance with prescribed rules but did not see the performance but only to see if the money spent already by the rules. From the above table, it can be made model of panel data analysis of the factors affecting economic growth in every regencies/cities in Java and Sumatra are interpreted as follows: In the model estimation above, it appears that the effect of variable crosssection that is different in every county and city in the region of Java and Sumatra to economic growth in every county and city where the district of Aceh Besar, Aceh Tengah, Dairi, Pasaman, Payakumbuh, Sawahlunto, Bengkalis, OKU, South Bengkulu, Kaur, Kepahiyang, Sukabumi and Blitar influence cross-section effect (effect operational area) which is positive, ie each region has a coefficient between 0.08 up to While in the districts/cities that have the effect to the cross-section (the effects of operational areas) is negative, ie Bireuen District, Asahan, Deliserdang, Agam, Padang, Kampar, Merangin, Tebo, Musirawas, Muaraenim, OKI, North Bengkulu, Lampung, Central, Brebes, Magelang, Solo and Kulon Progo is between up to Of the respective districts/cities in Java and Sumatra, an area that has the most impact on economic growth are Blitar (3.56) whose economies are sustained by the agricultural sector, Sawahlunto (3.38) which is supported by the mining economy, agriculture and tourism sectors, and Payakumbuh District (2.01) whose economy is sustained trade and agriculture sectors. While Brebes (-3.26), City Deliserdang (-2.433) and Central Lampung Regency (-2.08), which has the effect of a very small cross Section in economic growth. O. Conclusion From the results of data analysis can be summarized as follows: 1. Local Revenue had a positive effect on the economic growth of the region. The higher the PAD will have an impact on economic growth in the region. 2. Shopping for health has a positive effect on economic growth in the region, the greater the health budget greater economic growth created. The increase in the health budget impact on improving labor productivity, and the impact on improving people's welfare. P a g e 23