Bachelor Thesis
Impact of Employment Creation towards the Growth of GDP: Tourism Sector in Indonesia
Thesis of:
Adelita Asthasari Siregar 10418482
Supervisor:
D.H.J. Chen
Faculty of Economics and Business University of Amsterdam
Acknowledgement
For the strength and determination, every single breath and new day. Thank you for wonderful and countless opportunities to grow better in life. Thank you, Lord, Jesus, and Holy Spirit.
For my strongest fort, never ending support from miles away in Indonesia. For my dearest Maruli Tua Siregar, Dewi Sri Sotijaningsih, and Aprilia Asthasari Siregar. Terimakasih. Thesis ini Kakak persembahkan untuk Papa, Mama, dan Dek Lia. Untuk kesabarannya selama dua tahun kebelakang. Especially for every tear you shed, every prayer you sent. I love you.
Thank you for the best support system around: Mbak Henny dan Mbak Hana. I thank Mr. Joko Priyana, Ph.D for helping me tackles the grammatical errors, one of most horrifying parts in completing this thesis.
All of my dearest friends, for all funs and laughs along the process, thank you.
For my supervisor, Damiaan H.J. Chen, thank you for every suggestion, feedback, patient, and time given in helping me completes this bachelor thesis. Thank you.
And Amsterdam. Where everything was once a dream. Thank you.
Amsterdam, 21 July 2014 Adelita Asthasari Siregar
Content
Acknowledgement Contents ... i Abstract ... 1 Introduction ... 2 Literature review ... 5 Data Analysis ... 8 The Data ... 8 The Models ... 10 The Hypothesis ... 12 The Analysis ... 13Current Status of the Masterplan ... 18
Summary ... 19
Recommendation ... 20
Reference List ... ii
Appendix 1: Augmented Dickey Fuller Test ... vi
Abstract
Since 2011, Indonesia has been working on a 15-year master plan of the economic development. The master plan is a preliminary step for Indonesia to expand and accelerate economic development in order to support its transformation to a developed country. There are five main sectors included in the master plan: agriculture, mining, industry, tourism, and communication and transportation. Tourism is one of the recently growing sectors, with abundant resources and destinations. This study analyzed tourism impacts on the growth of Gross Domestic Product (GDP) and the result will be used to investigate the reasoning behind the inclusion of tourism in the master plan.
This study assumes that there is a positive impact of employment creation in tourism on the growth of GDP. In order to test the hypothesis, the study used a regression model analysis based on rearranged Okun’s law. Instead of seeing the effect of unemployment on GDP, the study focused on the impact of employment on GDP, especially the growth of GDP. The analysis used Ordinary Least Squares (OLS) regression involving 30-year monthly data ranged from 1984 to 2013 with the growth of GDP as the dependent variable and ten variables, i.e.: the growth of GDP from the previous year, interest rate, inflation rate, total employment, employment from five sectors in the master plan, and the rest of the employment as the independent. The analysis also included three dummy variables: d_98 for the 1998 crisis, d_IntRate for interaction between the dummy and the interest rate, and d_InflRate with the inflation rate.
The analysis found that the coefficient of employment in the tourism sector was 0.719. This finding shows that for every 1% in the tourism employment, the GDP growth increases by 0.007%. However, the statistic output for the tourism employment is not significant. In other words, statistically, employment creation in the tourism sector does not affect the growth of GDP.
Generating growth in GDP also creates social, economic, and environmental implications. Regulation and policy establishment are suggested to protect the society and the environment from the negative effect of the development. Increasing employment should also be followed by an increase in productivity. Further development in the infrastructure, the capacity building, and improvement in the transportation system is also expected to strengthen the productivity as an implication of employment creation.
Introduction
Looking at the GDP growth in the past 10 years, it is wise to say that economy in Indonesia is growing. Indonesia was able to maintain the positive growth during the crisis, along with China, while the rest of the world, such as the United States and countries in South East Asia, suffered negative growth.
There are several sectors that have been the primary support for the Indonesian economy. The agricultural sector is one of the main economic activities supporting millions of Indonesian people involved in this sector (Worldbank, n.d.). This sector shared Indonesia’s total GDP by around Rp330 trillion in 2004 and Rp1,311 trillion by 2013. In 1960’s crisis, the government found this sector as the main activity in rural areas that provided benefits to the economy; changes of the system took place to create a developmental and effective sector that was later used as one of the concentrated activities for a post-crisis treatment (Grabowski, 2011).
Figure 1. Growth of GDP of Indonesia compared to the world
source: data,worldbank.org (2014)
Another sector that supports the Indonesian economy and is still growing is the industrial sector, especially modern industrial techniques. The development of centralized industrial areas is one of the indicators that Indonesia pays attention to this sector’s growth. GDP contributed by the industrial sector exceeds the amount contributed by the agricultural sector. In 2004, the industrial sector’s contribution to Indonesia’s total GDP was around RP644 trillion and increased to Rp2,152 trillion in ten years’ time. Indeed, when a country started to earn higher incomes, the industrial sector tends to grow and shift the economic activity concentration of a developing country from the agricultural sector (Soubbotina, 2000).
In order to drift the economy faster and to exit from the rank of a low-income country, Indonesia established a 15-year master plan of the economic development, namely Masterplan for Acceleration and Expansion of Indonesia’s Economic Development1, that will work on 22 main economic activities. These activities can be categorized into five sectors: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1 Masterplan Percepatan dan Perluasan Pembangunan Ekonomi Indonesia
$5! 0! 5! 10! 15! 20! 2004! 2005! 2006! 2007! 2008! 2009! 2010! 2011! 2012! 2013! % "gr ow th "o f"G D P " China! Indonesia! Singapore! Thailand! United!States! world!
agriculture, industry, service, mining, and communication and transportation2. The master plan aims to make Indonesia as a high-income and developed country. To be categorized as a high-income country, its income per capita must be more than $12,616. Indonesia’s current income per capita is $3,420 (Worldbank, n.d.). Also, by 2025, Indonesia targeted the GDP to be ranged between $4-$4.5 trillion3, from the current GDP of $800 billions (Worldbank, n.d.) and the growth of GDP at 5.7% (Asian Development Bank, 2013). Expansion and acceleration effects are the main objectives of the master plan to help these main goals achieved.
To optimize the usage of the potential resources, the master plan works based on corridors. The 22 economic activities assigned to these corridors based on themes. The corridors and the themes are:
1. Sumatra as the central production and processing of natural resources and nation’s energy reservation,
2. The Java corridor supporting industrial and service activities, 3. Kalimantan as the center of mining and energy reservation,
4. Sulawesi as the central production and processing of agricultural, plantation, fishery, oil and gas, and mining products,
5. Bali-Nusa Tenggara as the tourism gateway and food reservation,
6. Papua-Kepulauan Maluku as the center of the national development of food, fishery, energy, and mining.
Besides the six corridors, the master plan will be strengthened by a higher degree of national connectivity and a better performance in science and technology. All three strategies will support the realization of the 22 economic activities, synchronization of national movement to revitalize the real sector performance, and the development of the center of excellence in each corridor of the economy.
This study is particularly interested in investigating the service sector. Soubbotina (2002) in a paper published by the Worldbank mentioned the expectation of a developed service industry in a high-income country and a growing industrial sector in a low-income country. The development in Indonesia’s service sector is therefore expected to follow the master plan. More specifically, this study is concerned with the tourism as a sector that is closely related to service and pronouncedly included in the master plan. The inclusion triggers curiosity on why this sector is chosen in the master plan to be developed and utilized. Since 2006, the number of incoming tourist has been continuing to rise, from four million to eight million in eight years’ time. This sector contribution to Indonesia’s total GDP is also significant, growing by 99% in 2013, reported by World Travel and Tourism Council in “Benchmarking Travel & Tourism in Indonesia”. The contribution of this sector to the total GDP also keeps increasing, reaching at Rp841 trillion by 2013. However, despite the growth, this sector’s share in the total GDP of Indonesia is smaller than other sectors’.
Natural disasters and national security events have also been a challenge in the development of the tourism sector in Indonesia. The tsunami in Nanggroe Aceh Darussalam !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2 Communication and transportation will be adopted as one sector due to data availability 3 kp3ei.go.id. KP3EI is a coordination committee for the implementation of the master plan
in 2004 essentially ruined most of the coastal tourism spots there (Fadillah, Dewi, and Hardjanto, 2012; Kelman, Spence, Palmer, Petal, and Saito, 2008). Bali bombing in 2010 also disturbed tourism activities that have been the core of the economy for years. The number of tourist arrivals in Bali declined after the incident and the government should pour USD200 million for tourism promotion in Bali (Pambudi, McCaughey, and Smyth, 2009). These facts show the vulnerability of this sector to external events.
The main intention of the master plan is to increase the pace of the GDP growth in order to achieve the goals using the development and expansion of economic activities. With its ups and downs, the impact of including the tourism sector therefore questionable and becomes the research question of this study. The study, using Okun’s law as the theoretical basis on the indicator of expansion, namely employment, tries to answer the research question. This paper assumes that tourism represents the whole service sector in the master plan. The output of this study is useful to explain the effectiveness of the established master plan and is expected to add the number of studies about the impact of expansion in the tourism sector through employment creation, how additional employment in this sector, known as labor intensive activity, affects the GDP growth, which is hard to be found.
Literature Review
The impact given by the tourism sector on the economy is well known through studies done by tourism organizations, such as World Travel and Tourism Council (WTTC) and United Nations World Tourism Organization (UNWTO), and scholars, such as ones that will be discussed throughout this section. The research focuses are varied from the impact of tourism on the economic growth (Ballaguer& Cantavella-Jorda, 2002) to the impact on employment (WTTC, 2012). However, the effect of employment creation in tourism on a country’s GDP and economic growth has not been popular among researchers. The numbet of studies discussing and focusing on this particular subject is limited. Most of the studies done in the tourism field partially discussed this matter, especially how tourism affects the employment and how tourism affects GDP. This study is important since expanding the economic sectors by creating more employment will sustain the economy, especially for a developing country, such as Indonesia, where unemployment is still problematic.
In 2013, based on annual research done by WTTC, the tourism contribution to the world‘s GDP grew by 3.1%, reaching the contribution to 9.5%, and outperformed other sectors in the economy such manufacturing, reportedly. Hence, the development and expansion in this sector should be managed in order to keep track of the positive impact it gives to the economy.
The development in this sector is particularly beneficial for a developing country and becomes one of the available options to further develop the economy (UNWTO). Balaguer and Cantavella-Jorda (2002) state that with the same amount of injection in this sector, a developing country would experience a bigger impact than if this action were done in developed countries, in line with the idea sustained by Toader (2011).
The effect of tourism on the economy comes from a number of channels. The first is through foreign exchange revenues. Foreign tourists’ consumption and their activities in a country is the main source of this benefit. Their expenditure will increase foreign exchange revenues (Dwyer & Forsyth, 1994; Elkan, 1975). The consumption will help the economy by acting as a source of incomes (Seetanah, 2011). Jenkins and Henry (1982) found that the importance of this channel towards a country’s economy could be seen through the degree of the government’s involvement. In other words, for a country where the foreign exchange from tourism highly affects the economy, the government will have a tendency to involve itself more in the tourism development.
Second, benefits in the development of tourism are also experienced by the investment sector. Iordache (2012) found that investment is triggered in the area where interest is high on the recreational and sports activities. In the case of Australia, Dwyer and Forsyth (1994) found that the possession of benefits from investment in tourism development, location of the development, and internalization of the investment are the determining issue for investing in tourism activities.
Via the investment sector, tourism then also acts as a better tool to drift the infrastructure development. Tourism activities generate incomes that provide opportunity for infrastructure to improve (Kreag, 2001). Availability and readiness of infrastructure, especially government-financed infrastructure, is an attractiveness determinant for tourism destination (Khadaroo and Seetanah, 2007). Enright and Newton (2004) in their works found
that the availability of infrastructure ranks high in determining a country’s competitiveness in tourism. The infrastructure development, therefore, gives a chance for tourism to develop further and continue to play roles in the growth of the economy.
Developing tourism infrastructure from scratch is not always needed. Ali, Huang, and Zhou (2011) found an ex-mining site used as a tourism destination in Yunnan, China. Digging activities opened the land surface and often exposed its inner layers. These layers create a geographical view that later stimulates tourism activities. They argue that the availability of the infrastructure from the previous mining activities, such as roads, can be re-utilized to support tourism activities. These benefits from another economic sector may increase the possibility of having a better-developed tourism sector with a lower cost for the infrastructure development.
Adding up possible benefits coming through tourism activity will create positive impacts on the GDP growth. Well-nurtured tourism will give a significant contribution to the national economy (Sugiyarto, Blake, & Sinclair, 2003).
Expansion in the tourism sector also provides benefits for the labor market. In 2010, tourism absorbed more than 235 million jobs worldwide, representing 8 percent of the global employment (Economic Review, 2011). The inability of the tourism sector to perform outsourcing is a channel to high employment absorption (Hjalager 2007). This sector also gives opportunity to people who are unfamiliar with the technology to access jobs (Hjalager, 2007), providing lower-middle occupations (Farver, 1984), and providing job opportunities with various skill requirements (Economics Review, 2011). This is an opportunity for a country where unemployment is still an issue and which has abundant tourism potential.
The impacts of tourism activities on the employment differ from one country to another. In Kenya and Tanzania, Elkan (1975) found that indeed tourism provides more jobs than the manufacturing sector. On the other hand, Elkan also found that the cost of employment in the tourism sector in both countries increased compared to that in another sector. In The Gambia, Farver (1984) found that an employer depends on expatriates for skilled jobs, such as chefs and managers, while local people with less formal education and mostly trained on-the-job are placed in unskilled positions, such as waiters and gardeners. Farver added that employment in tourism in The Gambia is mostly seasonal with 6 months off when the season is over.
In Indonesia, the tourism sector provides jobs in many layers, from formal to informal. The informal sectors are one of activity most likely present in tourist attractions in Indonesia, represented by street-vendors. The easy access, that the informal sector provides, attracts people to enter the market this way (Timothy and Wall, 1997).
In places such as Bali, where culture is still tightly held, employment causes social and economic implications. Cukier-Snow and Wall (1993) found that migration cause the possibility to shift the bond in banjar (traditional village). At the same time, migration also reduces the opportunity for local people to enter the job market. Culturally, Cukier-Snow and Wall quoting Picard (1990), the decision an employee should make between being committed to their job and attending a traditional ceremony threat the traditional bond among Balinese.
Indonesia should also be aware of the international migration. Referring to Hjalager (2007), it is naïve to hold the perception of the unchanged labor market, especially in low-income countries. Once the exposure from high-low-income countries enters the tourism sector,
the higher quality of service will be expected and people will start to expect the availability of skilled labor. Hjalager pointed that it will force the employee to gain a higher education and face the fact that there will be narrower opportunities to gain a job. The better policy regarding this was also suggested to avoid further suffering experienced by one party, namely labor.
When the job opportunities get narrower, unemployment becomes an issue. Okun’s law stated that when unemployment in a country increases, the GDP would decrease. Some tests show that the relationship is valid and that a negative correlation between unemployment and output exists. However, the magnitudes of the changes that occur are different across countries and sensitive to several factors, such as different methods used in the test.
Lee (2000) found that the countries investigated in the research exhibit different magnitudes of Okun’s law. Increased in unemployment caused a smaller lost in output in some countries compared to others. Country such as Japan and European countries are most likely to have a small coefficient of Okun’s law, due to their more rigid labor markets, while the US and Canada appear to have a bigger coefficient (Moosa, 1997).
Saget (2000) found that in transition countries, decrease in employment is bigger than decrease in the production. The study suggests that restriction on monetary and fiscal policies creates limitation on the demand side and becomes the cause of high unemployment. The outcome of Saget’s research showed that there is a correlation between GDP and employment in the observed countries, of which some show negative relationship.
Rearranging Okun’s law, the ability of tourism to absorb employment as found by scholars supposedly leads to higher GDP. Unfortunately, there are not many studies that have been done regarding this relationship. A closest study was done by Novianti (2013), who found that tourists’ consumption has a positive impact on tourism employment in Indonesia, both direct and total employment. Yet, this study still has not captured the impact of employment creation in tourism on the growth of GDP.
Data Analysis
To obtain a sufficient analysis, this study used one dependent variable, namely the growth of GDP (g), and ten independent variables: the growth of GDP from the previous year or the growth of GDPt-1 (gt1), interest rate (IntRate), inflation rate (InflRate), total employment (TOTEmp), agricultural employment (AGREmp), mining employment (MINEmp), industrial employment (INDEmp), tourism employment (TOUEmp), communication and transportation employment (CTEmp), and the rest of the employment (RESTEmp), also three dummy variables. The first dummy variable is d_98 indicating the period of economic crisis of 1998 in Indonesia. The second and the third dummy variables are the interaction between the dummy variable and two of independent variables: the interest rate (d_IntRate) and the inflation rate (d_InflRate). These variables were formulated into three models. OLS regression was the method to analyze the data. These models will be discussed later.
The Data
The analysis used 30-year monthly data, from 1984 to 2013. The data were obtained from three main sources: Datastream, World Travel and Tourism Council, and Badan Pusat Statistik (Indonesia Statistic Bureau). Monthly observations increase the size of data significantly compared to yearly data, from 330 to 3960 data. The bigger size of observations gives power to the data analysis.
The dependent variable used in this analysis was the growth of GDP. Denoted by g, the data obtained in the form of percentage and in constant price. The growth of GDP in constant price means that the growth has been adjusted to the inflation rate. This adjustment provides more accurate numbers compared to the unadjusted one.
The first independent variable is the growth of GDP from the previous year. The format of this variable is the same as dependent variable. This variable is also adjusted to the inflation. The second independent variable is the interest rate, obtained in percentages and adjusted to the inflation. The third independent variable is the inflation rate, obtained in percentage and based on CPI index. The government debt is the fourth independent variable. Obtained in percentage, the government debt used in this analysis is the percentage of debt to the nominal GDP. All variables mentioned above are some macroeconomics main indicators that have an unquestionable impact on GDP and the inclusion in the model will act as a basis for further model improvement. The dummy variables will indicate how the crisis in 1998 impacts the growth of GDP in Indonesia. These dummy variables equal to one in the period during crisis and else are zero. For the interaction, dummy variables will explain the additional impact given by the inflation and the interest rate during the period of crisis.
The rest of the independent variables are employments. Employment can be translated as the number of workers employed in the sector. It can also be assumed as labor hours employed (Gordon, 2010; Friedman and Wachter, 1974). This study used the number of workers working in each sector as the unit. Total employment depicted the sum of employment in Indonesia from all sectors. Employment from all sectors included in the master plan – agriculture, mining, industry, tourism, and communication and transportation – are the number of people employed in each sector. The rest of the employment, total!
employment!minus!the!sum!of!employment!in!the!sectors!tested, is the last independent variable in the analysis.
The theoretical basis for using employment as the expansion indicator in this study is Okun’s law. The law emphasizes on the negative relationship between unemployment and the GDP of a country. By the rearrangement of this law, a positive relationship between employment and GDP is expected.
The raw data of each variable was tested using Augmented Dickey Fuller (ADF) test to check the presence of a non-stationarity. It should be noticed since non-stationarity – the condition where the data show the occurrence of trend – causes mean, covariance, and correlation to be meaningless and the identification and estimation method not work (Ooms, 2004).
Table 1. Descriptive Statistic of Raw Data from the Variables
variable( Obs( Mean( Std.(Dev.( Min( Max(
g" 360( 4.986( 3.996( 917.929( 10.743( gt1" 360( 4.984( 4.002( 917.929( 10.743( IntRate( 360( 12.573( 9( 5.396( 70.81( InflRate( 360( 10.113( 11.644( 91.1( 82( Gdebt( 360( 37.002( 9.635( 23.908( 71.833( TOTEmp( 360( 8.59x107( 1.46x107( 6.08x107( 1.11x108( AGREmp( 360( 3.35x107( 6.6x106( 2x106( 4.16x107( MINEmp( 360( 9.6x105( 2.3x105( 5.79x104( 1.60x106( INDEmp( 360( 1.08x107( 2.26x106( 6.46x105( 1.54x107( TOUEmp( 360( 8.05x106( 1.84x106( 4.35x105( 1.09x107( CTEmp( 360( 4.6x106( 9.47x105( 2.77x105( 6.179x106( RESTEmp( 360( 2.8x107( 6.22x106( 2.12x107( 5.74x107(
The Models
Summarizing the ADF test4, the raw data of each variable, except for g, gt1, InflRate, Gdebt and RESTEmp, exhibited a non-stationarity. The data need to be transformed in order to eliminate the effect of the non-stationarity. The sectors employment data was transformed using natural logarithm, including the RESTEmp variable5. For IntRate, the result of the ADF test was ignored. The raw data of this variable were used in the analysis. The transformed data were tested again using ADF test and the outcome showed that the transformation stationed the data. The transformed data was used as the parameters in the regression replacing the original data.
To answer the research question, this study used three models. The first model included the growth of GDP from the previous year, inflation rate, interest rate, government debt, and three dummy variables. This equation aimed to provide a good basis to determine whether an improved model with employments gave a better explanation. The model is formally written as:
gt = β0 + β1 gt-1 + β2 IntRatet + β3 InflRatet + β4 Gdebtt + β5 d_98 + β6 d_IntRatet +
β7 d_InflRatet + εt,
with:
gt : the growth of GDP in current year,
gt-1 : the growth of GDP in previous year,
IntRatet : interest rate,
InflRatet : inflation rate (GDP deflator, percentage),
Gdebtt : government debt (percentage on total GDP, percentage),
d_98 : dummy variable for 1998 crisis,
d_IntRate : interaction between dummy variable for 1998 crisis and interest rate, d_InflRate : interaction between dummy variable for 1998 crisis and inflation rate. The total employment that was transformed using natural logarithm (ln_TOTt) added
to the second model in order to see the contribution of total employment in the growth of GDP. The second model is written as:
gt = β0 + β1 gt-1 + β2 IntRatet + β3 InflRatet + β4 Gdebtt + β5 d_98 + β6 d_IntRatet +
β7 d_InflRatet + β8 ln_TOTt + εt,
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
4 See appendix for the complete ADF test results 5 See appendix for reasoning
The last model used employment from five sectors that included in the master plan: agriculture, industry, mining, tourism, and communication and transportation. The rest of the employment in this model replaced the total employment. All employments variables transformed using natural logarithm and abbreviated using the first letter of the sector. The model, therefore, become:
gt = β0 + β1 gt-1 + β2 IntRatet + β3 InflRatet + β4 Gdebtt + β5 d_98 + β6 d_IntRatet +
β7 d_InflRatet + β8 ln_RESTt + β9 ln_At + β10 ln_Mt + β11 ln_It + β12 ln_Tt +
β15 ln_CTt + εt,
with:
ln_RESTt : Employment in activity outside five sectors in the model,
ln_At : Employment in agricultural activity,
ln_Mt : Employment in mining activity,
ln_It : Employment in industrial activity,
ln_Tt : Employment in tourism activity, and
The Hypothesis
Tourism participation to play important roles in Indonesia’s long-term economy is expected; this sector survived the crisis during 1997-1998 and was able to maintain the high growth (Sugiyarto et al., 2003). It is also known as one of the sectors that trigger a positive growth in the economy through several economic channels (Schubert et al., 2011).
Therefore, the coefficient of tourism employment sector in this study is expected to be positive. Although a positive coefficient will not be surprising, this study and output of the analysis will still be important to draw a conclusion whether employment creation in tourism helps the master plan to achieve its goals.
Improvement from the first model to the second and third models is expected, using adjusted R2 as the indicator.
The Analysis
The results of the regression models regression are presented in the table below.
(1) (2) (3)
VARIABLES Model 1 Model 2 Model 3
gt1 0.731*** 0.732*** 0.717*** (0.0254) (0.0255) (0.0262) IntRate 0.0256* 0.0269** 0.0184 (0.0131) (0.0137) (0.0162) InflRate -0.0147 -0.0151 -0.0181* (0.00940) (0.00948) (0.00968) Gdebt -0.0207*** -0.0205*** -0.0319*** (0.00583) (0.00586) (0.00716) d_98 -4.548*** -4.539*** -4.648*** (0.895) (0.897) (0.916) d_IntRate -0.107*** -0.108*** -0.0984*** (0.0331) (0.0334) (0.0346) d_InflRate 0.0652** 0.0657** 0.0628** (0.0293) (0.0293) (0.0293) ln_REST -0.145 (0.325) ln_A 9.158** (3.737) ln_M 0.189 (1.746) ln_I -3.487 (2.945) ln_T 0.721 (0.584) ln_CT -6.469** (2.906) ln_TOT 0.101 (0.317) Constant 2.140*** 0.269 -11.75 (0.314) (5.852) (10.04) Observations 359 359 359 R2 0.942 0.942 0.943 Adjusted R2 0.9407 0.9405 0.9411
Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
The first column on the table shows the first model that excludes employment. The R2 for this model is 94.2%, showing that 94.2% of the variance of the model is explained by the data. For the first independent variable, 1% increase in the growth of GDP from previous year supports the growth to increase by 0.731%.
The interest rate has a smaller impact on the growth of GDP. For 1% increases in the interest rate, the growth of GDP changes by 0.026%. A positive effect, although small, may grow smaller or even negative when the interest rate becomes higher. This is due to its relationship to GDP based on the macroeconomic theory. The higher interest rate will reduce the willingness to invest and consume which will lower the GDP and slow the growth down (Aron and Muellbauer, 2011; Taylor, 2000). The output is particularly strange since it against the nature of the interest rate and the GDP growth’s relationship. Hansen and Seshadri (2013) argued that this positive causality is possibly caused by the lag between the issuance of the policy and the reaction in the economy. It is possible, for the case of Indonesia, that the output from the high productivity period offset the cost of increasing the interest rate.
The inflation rate and the government debt have a negative impact on the growth of GDP. These two variables slowed the growth by 0.0147% and 0.0207% for 1% increase in each variable, respectively. Reinhart and Rogoff (2010), in their working paper, found that this relationship also happened in 44 countries during a crisis. A higher debt to GDP ratio indeed caused a slower growth. They also mentioned that the debt threshold that caused a slow-growth in GDP was lower for the emerging market. Indonesia should be aware of this threshold in order to avoid a slower growth.
In this model, it is to be noticed that the inflation rate is not significant. This may due to the inability of the data to give an explanation in the first model. However, it is better to further analyze this variable in the next two models.
In the second model, the R2 is not changed. Slight differences, however, occurred in the independent variables. The effect of previous year GDP on the growth of GDP is 0.732%. It means that in the second model, 1% increase in the growth of GDP has a 0.001% bigger effect on the current growth of GDP compared to the first model. The impact of a percentage increases in the interest rate on the growth of GDP increases by 0.001% in the second model to 0.027%. The inflation rate in this model has a bigger negative impact towards the GDP growth: 0.0151% deduction on the growth for every 1% increases in the rate. The government debt for this model shows a negative impact towards the growth of GDP in this model. For every percentage increases in this variable, the growth of GDP slowed by 0.0205%. The inflation rate is still insignificant in this model while the growth of GDP in the previous year and government debt are both significant at 1%.
The total employment (TOTEmp) as new a variable introduced in the second model provides a positive effect on the growth of GDP. The growth of GDP increases by 0.001% for every 1% increases in Indonesia’s total employment. Whether this percentage is good enough or not cannot be concluded. It may depend on the country’s economy; as stated by Lee (2000). Chamberlin (2011) argued that looking at a short term changes in the employment is still, what Chamberlin refer as, “interesting” (2011: 131) due to the closeness of the present and the past condition of the economy. Chamberlin added, for a longer observation term, employment imposed “by both cyclical and structural effects which has reduced the significance of Okun’s Law as a forecasting rule of thumb” (2011:131). The total
employment coefficient in this model, however, is not significant. This outcome showed that an increase in the total employment in Indonesia, statistically, could not give enough impact towards the growth of the GDP.
In the third model, the effect of a percentage change in the previous year’s growth of GDP decreases to 0.717% and simultaneously the interest rate that affects growth also decreases to 0.0184%. The inflation rate coefficient for the third model is still negative. For a percentage change in the inflation rate, the growth of GDP decreases by 0.0181%. The government debt’s coefficient in this model is the most negative among three models and affects the growth of GDP negatively by 0.03179%. R2 in the third model is 94.32%, slightly higher than the first two models. In this model, the inflation rate coefficient is now significant at 10%. On the other hands, the interest rate coefficient becomes insignificant in the third model. This outcome may give us information that the inflation rate explains the growth of GDP better when employment is present while the interest rate is the opposite.
To improve the analysis, five master plan sectors and the rest of Indonesia’s employment introduced in this model. The employment in the agricultural sector has the biggest coefficient compared to the other sectors. For every 1% increases in this sector’s employment, the growth of GDP will increase by 0.0916%. It may partially prove that although the agricultural area has been constantly decreasing in past years (Ministry of Agriculture of Republic of Indonesia, 2012), the output of this sector is still a major contributor in the growth of GDP. The coefficient is significant at 5%.
The industrial sector as one of the major contributors in the total GDP shows a negative impact on the growth of GDP in this model. The expansion in employment in this sector slows the growth of GDP by 0.035%. Referring to the concept of full employment, where the efficiency of the production is already achieved by the amount of labor provided, the industrial sector in Indonesia is probably already there. Employment creation in this sector therefore creates inefficiency. This sector coefficient is surprisingly not significant, considering its contribution to the nominal GDP and the growth of GDP, amounting at 1.62% in 2013 (The Jakarta Post, 2013). A reasonable explanation for the insignificant coefficient is the range of the observation’s period. Since this sector’s development is relatively new in Indonesia, significant impact most likely to be captured if the analysis used a narrower and a more recent range of time.
Employment creation in the mining sector results in a negative impact on the growth of GDP. For an additional percentage of employment in the mining sector, the growth of GDP is lowered by 0.002%. The characteristic of the exhaustible natural resources, the main commodities of the mining sector, may lead to inefficiency for every one more person employed in this sector. The insignificance of this sector’s coefficient in this model may be assumed as a very small contribution to the growth since assuming zero contribution seems irrational6. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 6!Insignificance!should!not!be!directly!taken!as!an!indicator!of!zero!contribution.!Peterson!(n.d.)!argued! that!inclusion!of!a!variable!in!a!model!must!be!based!on!consideration!that!it!will!give!a!contribution!to! explain!the!model.!Assuming!that!nothing!is!contributed!by!an!insignificant!coefficient!therefore!is!not! wise.!
Employment creation in the communication and transportation sector in Indonesia also causes a slower growth in GDP. The coefficient shows an expansion in this sector lowers the growth by 0.065%. The close relationship between the communication and technology sector becomes a weakness for Indonesia. Technology sector in Indonesia is underdeveloped, with the amount of investment only 0.08% of GDP (Worldbank, n.d.). Adding more people to work in this sector without considering the possible development will trigger inefficiency. The master plan mentioned the development in science and technology as one of the pillars. A significant amount of investment should be placed to boost the development. From the transportation sector, an insufficient infrastructure development (Worldbank) becomes a plausible reason of employment inefficiency. This sector coefficient is significant at 5%.
Tourism as the center of this study shows a positive coefficient. For a percentage increase of employment in this sector increases the growth by 0.007%. The insignificance of the coefficient of the tourism employment in the model, however, should be taken as statistically not contributing to the growth of GDP. In other words, employment creation in tourism statistically will not help the master plan achieving its goals.
This output should place a warning for the master plan committee. Expanding and developing the tourism sector should be done carefully so that it will not lead to the direction that will weaken the contribution of this sector to the growth of GDP. Employment creation in tourism, based on this study, is not a good direction to which the master plan should lead. This is may be due to the underdevelopment of this sector. Unregulated and uncontrolled periods of development (ILO, n.d.) can also be the cause behind a small contribution of the tourism sector in this study.
Attention should also be paid to the coefficient of the rest of the employment, ln_REST that shows a negative impact on the growth of GDP. For 1% increase in the employment of these sectors, the growth of GDP will decrease by 0.001%. Although this coefficient is not significant, it still gives information on sectors that is not beneficial to be expanded through employment creation. Studies in these sectors could provide insight of the prospects that may be useful for further development once the period of the master plan ends without jeopardizing the growth of GDP that already achieved.
One important parameter that should not be overlooked in analyzing the output is the adjusted R2. Since the analysis uses three different models and a variable is, or variables are, added from one model to the other, R2 is not a good parameter considering its value that will automatically increase as more variables added into the model. The adjusted R2 value, that will not automatically increase as more parameters added, since it depends on whether additional variables help to explain the model or not, therefore, will be a good indicator on how models are improved. In the first model, adjusted R2 is 94.07%7. For the second model, it decreases to 94.05%. A decrease in the value of the adjusted R2 indicates that an additional variable in the second model, total employment, is not able to explain the model better. It is not surprising knowing that the coefficient of the total employment is not significant. For the last model, adjusted R2 is 94.11%, higher than the first and, definitely, the second models. Here we can conclude that, indeed, the sector’s employment improves the model.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Analyzing the second and the third models further, it is obvious that the impact of employment creation on the GDP growth is better explained by a specific sector than a total. The second model, where total employment is used as the employment parameter, shows 0.001% contribution to the GDP growth and it is not statistically significant. While in the third model, where the employments are broken down into specific sectors, the coefficients are either bigger or smaller than 0.001%, and either significant or not. Taking employment in the agricultural sector as an example, the impact of employing 1% more in this sector increases the growth of GDP by more than 0.09% and significant at 5%. This percentage showed that there are sectors that still need additional human capital as part of its development, while there are sectors that do not need that, such communication and transportation that has negative and significant coefficient. Observing the impact of employment creation on the growth closer to each sector will provide more information.
The discussion on the growth of GDP will not be complete without discussing the hurdle to achieve it. One of the most prominent obstacles is an economic crisis. Indonesia, in 1998, experienced an agitating economic downturn. The crisis affected the whole system in Indonesia: economically, socially, and culturally. This study encountered spikes in the data around the period of the crisis and captured it as dummy variables. In all of three models, these dummy variables are significant at 1% for d_98 and 5% for d_IntRate and d_InflRate. During the crisis, the economy in Indonesia grew slower by approximately 4.5%, all models shows in the coefficient of d_98. When the analysis paired the dummy variable with two macroeconomics indicators used in the models, the interest and the inflation rate, the output shows how each of it behaved in affecting the growth of GDP during the crisis.
In the period of the crisis, one percent increase in the interest rate cost the growth of GDP a 0.107% deduction, shows in the d_IntRate coefficient. As it was discussed before, a higher interest rate will lower the willingness to invest and spend; both activities are the main components of the GDP. During crisis, this behavior amplified since people are more reluctant to gamble using their money. Inflation rate, on the other hand, will trigger more production. Also during the 1998 crisis, Indonesia’s growth of GDP increase by around 0.06% per 1% increase in the inflation rate shows in the d_InflRate coefficient.
Current Status of the Master Plan
The master plan has only been going for the past three years. It may be immature to conclude where the plan may end up. By far, based on the report published by KP3EI (2013), the validation process made the corridors lose its projects in the real sector and gain some in the infrastructure sector. Table 2 shows the details of the value of the projects. The percentages of the groundbreaking projects are still relatively low with the lowest achievement of 9.86% in corridor Sulawesi and the highest realization of 27.58% in corridor Kalimantan. With eleven years left, Indonesia still needs to catch up with the plans while not neglecting the possibility to deal with issues that arise in between.
In realizing the plan, every corridor is facing common issues. Regulations and licensing still become a major drawback that cost continuation of several projects is on hold. Several regulations that already passed the legalization process are ready to be implemented. Specifically for Java, as the center of industrial activities, logistic becomes the main limitation in the process of execution.
Unavailability of human resources is also one of the major issues that slow the pace of the master plan. This issue is closely related to the limited number of learning centers that could provide competent and skilled labor. Professional and skilled labor that is directly related to the economic activity is anticipated to be actively involved to avoid idle time in waiting for labor that still in the training process.
Lack of the infrastructure availability also holds activities back. This may be the reason why some real sector’s projects did not pass the validation; support from infrastructure is possibly not available for these activities. Increase in the number of validated projects in infrastructure may explain the effort to start realizing projects in the real sector.
Sumatera Jawa Kalimantan Sulawesi Bali-Nusa
Tenggara
Papua-Kepulauan Maluku
Real Sector Plan 444 1290 869 243 143 448 Validated 404 323 281 98 170 464
Infrastructure Sector Plan 414 799 167 111 67 162
Validated 423 923 165 186 70 121 Groundbreaking Projects 107 210 123 28 43 82 Groundbreaking Projects % 12.94% 16.85% 27.58% 9.86% 17.92% 14.02%
Table 3. Current investment status, all in Trillion Rupiah Source: Implementation Report, KP3EI, 2013 (kp3ei.go.id)
Summary
Despite obstacles and issues that arise in between the master plan realization, Indonesia is left with eleven years of opportunity to achieve its goal. The utilization of sectors that prominently contributes to the economy will help to realize the goals. The vailability of the infrastructure and the human capital should be established in order to support the economic activities.
The results of the data analysis show that total employment has no clear cut on whether it has a positive or negative impact to the growth of GDP due to an insignificant coefficient. The mining and industrial sector also has the same issue. Meanwhile, the agricultural sector shows a prominent impact, offsetting other sectors by giving a positive contribution of 0.092% and employment creation in the communication and transportation sector affects the growth of GDP negatively, both coefficients are statistically significant.
The main focus of this study, tourism, shows a positive coefficient. However, this coefficient is not statistically significant. This outcome indicates a very small or even zero impact of tourism employment creation on the growth of GDP. The master plan development and expansion agenda hereby should be directed towards different focuses besides employment creation.
As the analysis goes further, the outcomes of the second and the third models obviously show that total employment could not explain the whole story of the impact of employment creation on the growth since a closer look at the specific sector employment will give more opportunity to exploit the benefit. The government should also anticipate the presence of an economic crisis since this event quantitatively showed a significant negative impact on the growth of GDP.
It needs to be noted that the contribution of all parameters is a percentage effect on the percentage of the growth of GDP, not on the total nominal amount. Hereby, small percentages may represent significant portions of a nominal amount of GDP.
The limitation of this study is that the models are simplified to explain the impact of employment creation on the growth. In the real economy, this simplification may not apply. Looking at the value of R2 and adjusted R2 raise indication that there are other variables that will suit these models in an attempt to explain the GDP growth better. And last, as it was mentioned in the introduction, this study assumed that service industry was only represented by the tourism sector. This oversimplified assumption ignores the possible service activities that arise in other sectors.
Recommendation
The master plan seems like an ambitious aim as it is referred to by Strategic Asia that also proposed nine major barriers of this master plan: a low degree of socialization, unclear line between the master plan and RPJMN (long-term national development plan), needs of reformation in the institutional and regulation bodies, a development in the infrastructure, the discrepancies between regions, human resources that are still underdeveloped, financing problems, and the goals that may conflict the maintenance of sustainability. Each corridor has possibility to face these barriers in different circumstances and regulators should be able to address these issues with a policy that fits each corridor well8.
From the output, the previous year’s growth of GDP showed a positive impact on the growth of the following year. Therefore, it is better if the growth of GDP is managed in order to keep the benefit. For other variables that have negative impact on the growth of GDP, regulations will be needed in order to guide the movement so that it will not worsen the economy.
Employment creation in each sector should also be regulated. Paying attention to sectors that has positive impacts on employment absorption will maintain effectiveness and efficiency in obtaining a higher growth. Law regulations of the sectors that allow foreign investment are needed to protect the rights of Indonesian people. A downside effect that follows the expansion should be avoided so that it will not offset the goodness brought by the development9. The welfare of people exposed to this development and expansion also needs to be considered. The sustainability of the nature and the environment cannot be neglected either. Balance is needed in order to create a long-term positive effect from this Materplan.
An increasing amount of labor employed in the sectors showing a positive impact will be better if followed by a higher degree of productivity. The development in infrastructure, along with higher integration in the transportation system, therefore, should not be overlooked. The development in human resources is also recommended. For a sector, such tourism, where human resources are one of the main components, skilled labor will be demanded in the future. Training and education are essential to create a competitive human capital.
Kwik Kian Gie, an Indonesian economist, argues10 that the development that focuses on the particular strength as used in the master plan is considered as a conventional approach. However, he suggests, that it is still better than neglecting the potential resources. Regarding this thought, an advanced approach to optimize any value-added resources possessed by Indonesia is expected. Hopefully, through a better technique in research, technology, and science that will be available following the 15 years’ implementation of the master plan, Indonesia will be able to utilize other welfare-generator sources for a better self. As the Bible story once said, for whichever talent or wealth given is not wisely used, it will be a waste (Matthew 25:14-30).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
8 Taylor-made policy will give solution with higher precision and allow faster settlement.
9 Indonesia should be well aware of the presence of immiserising development. Proper calculation and
consideration will avoid worsened welfare of Indonesian people.
10 Gie also discussed his proposal on a master plan in urban planning to increase wellbeing of Indonesian people.
He argued that the government should be involved more in urban and social economy planning to avoid losing potential sources for economy to private or even foreign investors.
References list
Aron, J. and Muellbauer, J. “Interest Rate Effects on Output: Effidence from a GDP Forecasting Model for South Africa.” IMF Staff Paper 49, (2002) : 185-213. Balaguer, J. and Cantavella-Jorda, M. “TOURISM AS A LONG-RUN ECONOMIC
GROWTH FACTOR: THE SPANISH CASE”. Applied Economics 34, no. 7, (2000) : 877-884.
Blake, A., Sinclair, M. T., and Sugiyarto, G. “TOURISM AND GLOBALIZATION Economic Impact in Indonesia.” Annals of Tourism Research 30, no. 3 (2003) : 683-701. doi:10.1016/S0160-7383(03)00048-3.
Chang, C. and Lee, C. “Tourism development and economic growth: A closer look at panels.” Tourism Management 29, (2007) : 180–192. doi:10.1016/j.tourman.2007.02.013.
Chamberlin, G. “Okun’s Law revisited”. Economic and Labour Market Review, (2011) : 104-132.
Cukier-Snow, J. and Wall, G. “Tourism Employment: Perspective from Bali.” Tourism Management 14, no. 3 (1993) : 195-201.
Elkan, W. “The Relation between Tourism and Employment in Kenya and Tanzania.” Journal of Development Studies (1975) : 123-130.
Fadillah, A., Dewi, T. G., and Hardjanto, A. “Analysis of Alternative Strategy in Coastal Tourism Developmetn in Aceh Besar, Indonesia after Tsunami Disaster.” International Journal of Social Science and Humanity 2, no. 3 (2012) : 206-212.
Farver, J.A. “ Tourism and Employment in Gambia.” Annals of Tourism Research 11, (1984) : 249-265.
Friedman, B.M. and Wachterm M.L. “Unemployment: Okun’s Law, Labor Force, and Productivity.” The Review of Economics and Statistic 56, no. 2 (1974) : 167-176.
Gordon, R.J. “Revisiting and Rethinking the Business Cycle Okun’s Law and Productivity Innovations”. American Economic Review: Papers & Proceedings 100 (2010) : 11-15.
Hansen, Bruce E., and Seshadri, A. “Uncovering the Relationship between Real Interest Rates and Economic Growth.” Ann Arbor MI: University of Michigan Retirement Research Center (MRRC) Working Paper, WP 2013-303. http://www.mrrc.isr.umich.edu/publications/papers/pdf/wp303.pdf
Hjalager, A. “Stages in the Economics Globalization of Tourism.” Annals of Tourism Research 34, no. 2 (2007) : 437-457.
Huang, G., Xhou, W,. and Ali, S. “Spatial Patterns and Economic Contributions fof Mining and Tourism in Biodiversity Hotspots: A Case Study in China.” Ecological Economics 70, (2011) : 1492-1498.
Kelman, I., Spence, R., Palmer, J., Petal, M., and Saito, K. “Tourists and disaster: lessons from the 26 December tsunamis.” Journal of Coastal Conservation 12, (2008) : 105-113.
King, B., Pizam, A., and Milman, A. “SOCIAL IMPACTS OF TOURISM Host Perceptions. Annals of Tourism Research 20, (1993) : 650-665.
Komite Percepatan dan Perluasan Pembangunan Ekonomi Indonesia. Laporan Perkembangan Pelaksanaan MP3EI. Indonesia. 2013.
Matthew 25:14-30 (New Living Translation)
McKercher, B., Law, R., and Lam, T. “Rating Tourism and Hospitality Journals.”
Tourism Management 27, (2006) : 1235-1252.
doi:10.1016/j.tourman.2005.06.008.
Novianti, J. “Tourism Consumption in Relation to Employment: the Case of Indonesia.” (Bachelor Thesis, University of Amsterdam, 2013)
Ooms, M. “Linear Time Series Models for NonStationary data.” Econometrics II – Chapter 7.3, Heij et al (2004).
Pambudi, D., McCaughey, N., and Smyth, R. “ Computable general equilibrium estimates of the impact of the Bali bombing on the Indonesian economy.” Tourism Management 30, (2009) : 232-239.
Ross, A.K. and Spiegel, M.M. “Cross-coutry causes and consequences of the of the crisis: An update.” European Economic Review 55, (2011) : 309-324.
Schluter, R. and Turgut, V. “Resident Attitudes Towards Tourism in Argentina.” Annals of Tourism Research 15, no.3 (1988) : 442-445.
Schubert, S., Brida, J. G., and Risso, W. A. “The impacts of international tourism demand on economic growth of small economies dependent on tourism.” Tourism Management 32, (2011) : 377-385.
Seetanah, B. “ASSESSING THE DYNAMIC ECONOMIC IMPACT OF TOURISM FOR ISLAND ECONOMIES.” Annals of Tourism Research 38, no. 1 (2011) : 291–308.
Sofiels, T. H. B. “Indonesia’s National Tourism Development Plan.” Annals of Tourism Research 22, no. 3 (1995) : 690-694.
Strategic Asia. “Implementing Indonesia’s Economic Master Plan (MP3EI): Challenges, Limitations, and Corridor Specific Differences.” 2012.
Taylor, J.B. “Teaching Modern Macroeconomics at the Principles Level.” American Economic Review 90, no. 2 (2000) : 90-94.
Timothy, D. J. “PARTICIPATORY PLANNING A View of Tourism in Indonesia.” Annals of Tourism Research 26, no.2 (1999) : 371-391.
Timothy, D.J. and Wall,G. “SELLING TO TOURISTS Indonesian Street Vendors.” Annals of Tourism Research 24, no. 2 (1997) : 322-340.
Toader, V. “The Economic Effects of Tourism: the Case of Romanian Economy.” Studia Universitatis Babes Bolyai-Negotia 1, (2011) : 53-66.
E-source
“Economy.” www.adb.org. Accessed on July 15, 2014.
http://www.adb.org/countries/indonesia/economy
“Employment in the Tourism Industry to Grow Significantly.” Economic Review. Last modified May, 2011.
http://web.b.ebscohost.com.proxy.uba.uva.nl:2048/ehost/pdfviewer/pdfviewer
?sid=8f8055ba-5ea9-4858-9f6f-757f70f339f3%40sessionmgr115&vid=2&hid=121
Gie, Kwik Kian. “Titik Lemah Program MP3EI.” Radio interview, unknown time and place. https://archive.org/details/KwikKianGie-titikLemahProgramMp3ei Hadwerdoyo, C.H. “Indonesia: An Industry-led Growth Economy.” The Jakarta Post.
Last modified January 4, 2013.
http://www.thejakartapost.com/news/2013/01/04/indonesia-an-industry-led-growth-economy.html
Peterson, S.P. “Interpreting P-Values.” Accessed on July 14, 2014. http://users.vcu.org/~sppeters/econ641/
“Rural & Agriculture in Indonesia.” Web.worldbank.org. Accessed on July 11, 2014. http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/EASTASIAP ACIFICEXT/0,,contentMDK:23186152~pagePK:146736~piPK:146830~theSi tePK:226301,00.html
“Sustainable Tourism in Indonesia.” www.ilo.org . Accessed on July 16, 2014.
http://www.ilo.org/sector/activities/projects/WCMS_159014/lang--en/index.html
“Transport in Indonesia.” Web.worldbank.org. Accessed on June 19, 2014.
http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/EASTASIAP ACIFICEXT/EXTEAPREGTOPTRANSPORT/0,,contentMDK:20458729~m enuPK:2066318~pagePK:34004173~piPK:34003707~theSitePK:574066,00.ht ml
“World Bank Supports Move to Accelerate Indonesia’s Acceleration Growth through Science, Technology, and Innovation.” Worldbank.org.
Last modified March 29, 2013.
http://www.worldbank.org/en/news/press-release/2013/03/29/world-bank- supports-move-accelerate-indonesia-economic-growth-science-technology-innovation
Data List
Badan Pusat Statistik. Jumlah Angkatan Kerja, Penduduk Bekerja, Pengangguran,
TPAK dan TPT, 1986–2013.
http://bps.go.id/tab_sub/view.php?kat=1&tabel=1&daftar=1&id_subyek=06 ¬ab=5
Badan Pusat Statistik. Penduduk 15 Tahun Ke Atas yang Bekerja menurut Lapangan
Pekerjaan Utama 2004 – 2013.
http://bps.go.id/tab_sub/view.php?kat=1&tabel=1&daftar=1&id_subyek=06 ¬ab=2
Badan Pusat Statistik. Produk Domestik Bruto Atas Dasar Harga Berlaku Menurut
Lapangan Usaha (Miliar Rupiah), 2004-2013.
http://bps.go.id/tab_sub/view.php?kat=2&tabel=1&daftar=1&id_subyek=11 ¬ab=1
World Travel and Tourism Council. Indonesia: Travel and Tourism Total Contribution to Employment. wttc.org
World Travel and Tourism Council. Indonesia: Capital Investment. wttc.org Worldbank. International tourism, number of arrivals. data.worldbank.org
Appendix 1: Augmented Dickey Fuller Test
As it was discussed, occurrence of non-stationarity in the data is possible and this presence should be noticed. Augmented Dickey Fuller test performed to the original data to identify this issue. Employment from each sector and the other five variables (previous year growth, inflation rate, interest rate, and government debt) will be tested. Hypothesis for this test are the presence of trend as H0 and the absence of it as H1. H0 will be rejected if the test statistic value is less than critical value; here we are using 95% confidence level. The result of preliminary test using original data showed below.
Variables Test statistic Critical value 5% -2.876 g -2.926 IntRate -2.136 InflRate -4.99 Gdebt -3.17 TOTEmp -0.209 RESTEmp -6.042 AGREmp -2.091 MINEmp -0.378 INDEmp -0.267 TOUEmp -1.858 CTEmp -1.944
Outputs show the presence of trend in IntRate and all employment variables except rest of employment. It is necessary to transform the data and to take another Augmented Dickey-Fuller test check whether the transformation stationed the data. The results are as follow.
Variables Test statistic Critical value 5%
-2.876 ln_TOT -6.752 ln_REST -3.735 ln_A -19.382 ln_M -19.097 ln_I -19.567 ln_T -19.026 ln_CT -19.139
Table 5. Output of ADF test, transformed data
For interest rate, this study will ignore the ADF test result. No transformation was applied to IntRate. Instead, this variable was analyzed further using dummy interactions. Meanwhile, for employments, natural logarithm is the most sensible way to interpret the changes. Adding one person into sectors may not give much information on the impact on the growth of GDP. Using the natural logarithm means the output of the analysis is percentage changes in employment. The output will give a more sensible number to be interpreted.
In the second table also included test for RESTEmp that previously not having non-stationary issue. The reason in adjusting the RESTEmp variable is to produce a similar way of interpreting the output as the other employment variable.
Appendix 2: Regression Output of all Models
Table 6. Output Regression Model 1
Table 7. Output Regression Model 2
_cons 2.140011 .3136254 6.82 0.000 1.52319 2.756832 d_InflRate .0652385 .0292615 2.23 0.026 .0076886 .1227884 d_IntRate -.1069337 .0331108 -3.23 0.001 -.1720543 -.0418131 d_98 -4.548247 .8949945 -5.08 0.000 -6.308473 -2.788021 Gdebt -.0206854 .0058263 -3.55 0.000 -.0321442 -.0092265 InflRate -.0147119 .0094023 -1.56 0.119 -.0332037 .00378 IntRate .0256085 .013059 1.96 0.051 -.0000752 .0512922 gt1 .73123 .0253558 28.84 0.000 .6813615 .7810985 g Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 5733.5202 358 16.0154195 Root MSE = .97459
Adj R-squared = 0.9407
Residual 333.38652 351 .949819144 R-squared = 0.9419 Model 5400.13368 7 771.447668 Prob > F = 0.0000
F( 7, 351) = 812.20
Source SS df MS Number of obs = 359 . reg g gt1 IntRate InflRate Gdebt d_98 d_IntRate d_InflRate
. reg g gt1 IntRate InflRate Gdebt d_98 d_IntRate d_InflRate ln_RESTEmp ln_A ln_M ln_I ln
_cons .2688526 5.851646 0.05 0.963 -11.23996 11.77766 ln_TOT .1013823 .3165946 0.32 0.749 -.5212849 .7240496 d_InflRate .0657284 .0293389 2.24 0.026 .0080257 .1234311 d_IntRate -.1082097 .0333918 -3.24 0.001 -.1738836 -.0425358 d_98 -4.539377 .8965688 -5.06 0.000 -6.302717 -2.776037 Gdebt -.0205202 .0058565 -3.50 0.001 -.0320386 -.0090018 InflRate -.015062 .0094776 -1.59 0.113 -.0337023 .0035782 IntRate .0268746 .0136603 1.97 0.050 7.86e-06 .0537412 gt1 .7317984 .0254503 28.75 0.000 .6817436 .7818531 g Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 5733.5202 358 16.0154195 Root MSE = .97583
Adj R-squared = 0.9405
Residual 333.28887 350 .952253915 R-squared = 0.9419 Model 5400.23133 8 675.028916 Prob > F = 0.0000
F( 8, 350) = 708.87
Source SS df MS Number of obs = 359 . reg g gt1 IntRate InflRate Gdebt d_98 d_IntRate d_InflRate ln_TOT
Table 8. Output Regression Model 3 " _cons -11.74785 10.04203 -1.17 0.243 -31.49917 8.003458 ln_CT -6.46946 2.906165 -2.23 0.027 -12.18549 -.7534284 ln_T .7205037 .5843477 1.23 0.218 -.4288287 1.869836 ln_I -3.486692 2.944836 -1.18 0.237 -9.278783 2.3054 ln_M .1893246 1.745874 0.11 0.914 -3.244572 3.623221 ln_A 9.157801 3.736922 2.45 0.015 1.807784 16.50782 ln_RESTEmp -.1453499 .3250151 -0.45 0.655 -.7846103 .4939105 d_InflRate .0628316 .0292975 2.14 0.033 .0052074 .1204558 d_IntRate -.0983833 .034552 -2.85 0.005 -.1663424 -.0304243 d_98 -4.648116 .9157145 -5.08 0.000 -6.449202 -2.84703 Gdebt -.0318583 .0071569 -4.45 0.000 -.045935 -.0177817 InflRate -.0180591 .0096757 -1.87 0.063 -.0370899 .0009716 IntRate .0183867 .0162186 1.13 0.258 -.0135131 .0502866 gt1 .7169984 .02622 27.35 0.000 .6654272 .7685696 g Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 5733.5202 358 16.0154195 Root MSE = .97125
Adj R-squared = 0.9411
Residual 325.446565 345 .943323377 R-squared = 0.9432 Model 5408.07363 13 416.005664 Prob > F = 0.0000
F( 13, 345) = 441.00
Source SS df MS Number of obs = 359 > _T ln_CT
