Dynamics between economic growth, labor, capital and natural resource abundance in Iran: An application of the combined cointegration approach

Resources Policy 49 (2016) 213–221

Contents lists available at ScienceDirect

Resources Policy journal homepage: www.elsevier.com/locate/resourpol

Dynamics between economic growth, labor, capital and natural resource abundance in Iran: An application of the combined cointegration approach Khalid Ahmed a,b,n, Mantu Kumar Mahalik c, Muhammad Shahbaz d,e a

University of Cambridge, Alison Richard Building, 7 West Road, CB3 9DT, Cambridge, UK Sukkur Institute of Business Administration, Airport Road, 65200 Sukkur, Pakistan c Department of Humanities and Social Sciences (DHSS), National Institute of Technology (NIT), Rourkela 769008, Odisha, India d Energy Research Center, COMSATS Institute of Information Technology, Lahore, Pakistan e Montpellier Business School, 2300 Avenue des Moulins, 34080 Montpellier, France b

art ic l e i nf o

a b s t r a c t

Article history: Received 5 January 2016 Received in revised form 4 June 2016 Accepted 6 June 2016

This paper discusses the missing case of Iran and tests the resource curse hypothesis using the updated time-series data over the extended period of 1965–2011. This study incorporates economic growth as a function of natural resources, exports, capital and labor in a Cobb–Douglas production function. The results of Bayer-Hanck combined cointegration test confirm that the underlying variables are cointegrated, while the finding from the long-run analysis validate the resource curse hypothesis and suggest that the natural resource impede economic growth in Iran. A 1% increase in natural resource production results in a 0.47% decline in GDP. This suggests that the exploitation of natural resources negatively affects the competitiveness of other sectors and limits their ability to contribute to economic growth. Furthermore, the results of the causal analysis conclude that there is a feedback effect between natural resource abundance and economic growth. These findings are useful for the development of policy controls in the case of Iran. & 2016 Elsevier Ltd. All rights reserved.

JEL: C3 O4 O5 Keywords: Natural resource abundance Economic growth Labor Capital Iran

1. Introduction The debate on whether natural resource abundance hinders economic growth started with the seminal study of Sachs and Warner (1995). Their study analysed a sample of 95 developing economies, and concluded that economies with a limited endowment of natural resources outpace the economic growth of resource abundant economies. These results led to the formulation of the ‘resource curse hypothesis’. Since then, the literature on natural resource abundance and the economic growth nexus has been testing the resource curse hypothesis on various macro- and micro-economic indicators, but the findings have been mixed (Collier and Goderis, 2008). Over the past two decades, there has n Corresponding author at: University of Cambridge, Alison Richard Building, 7 West Road, CB3 9DT Cambridge, UK E-mail addresses: [email protected], [email protected] (K. Ahmed), [email protected] (M.K. Mahalik), [email protected], [email protected] (M. Shahbaz).

http://dx.doi.org/10.1016/j.resourpol.2016.06.005 0301-4207/& 2016 Elsevier Ltd. All rights reserved.

been plenty of literature exploring the oil-rich economies, which have found a negative relationship between resource richness and economic growth (Ross, 1999; Sachs and Warner, 2001; Papyrakis and Gerlagh, 2004; Robinson et al. 2006). Amid several theoretical explanations presented in the literature, the most common is that resource rich economies have a tendency of resource dependence (Williams, 2011; Dubé and Polèse, 2015). However, Brunnschweiler and Bulte (2008) who refer to the resource curse hypothesis as a ‘red herring’, claim that resource abundance determines resource dependence, but does not affect the ratio of economic growth. Furthermore, the findings of Papyrakis and Gerlagh (2004) suggest that resource abundance only affects economic growth when considered in isolation. In addition to the mixed emperical evidence, a sufficient gap persists in the theoretical framework presented to date. For example, Van der Ploeg and Poelhekke (2009) consider volatility as the quintessential characteristic of the resource curse. Whereas, Frankel (2010) concludes that commodity exporting countries perform well and control the crowding out effect of resource endowment. James (2015) suggests that the

214

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

20.00

GDP growth (annual %)

15.00

20%

10.00 5.00 0.00 1960

7% 1970

1980

1990

2000

2010

2020

-5.00

73%

-10.00 -15.00 Year

Fig. 1. Iran’s GDP Growth Trend (1965–2011). Source: World Bank (2014).

resource curse depends on industry makeup, because it has little effect on non-resource sectors. In a nutshell, the debate on natural resource abundance and economic growth is still of great interest to both researchers and policy makers. This study attempts to contribute to the existing literature by examining the specific case of Iran. Despite having abundant natural resources and slow economic growth, the recent literature on the subject ignores the most likely case of Iran under the ‘resource curse hypothesis’. Moreover, the recent step of ending the long lasting economic sanctions on Iran further enhances the importance of this study and suggests concrete policy implications for the country's future resource and economic growth policies. Iran is a country abundant in rich mineral resources. It possesses the 2nd largest substantiated natural gas and 4th largest oil reservoir in the world (EIA, 2014). Iran is also the 2nd largest country both in terms of area and population in the MENA1 region. However, as shown in Fig. 1, the economic growth of Iran has been very slow, in fact negative in many periods. The average growth rate of the country during 1965–2011 was less than 3% and the linear trend has been downward sloping shown by the red and blue lines, respectively (see Fig. 1). Other macroeconomic indicators2 also reflect unhealthy performance, e.g. income per capita (4769 US$ in nominal terms), inflation rate (18.2% in CPI) and unemployment (10.4%), population below poverty line (0.7%). The Iranian economy is highly dependent on the hydrocarbon sector, which has limited the country's ability and diversified its economy. In 2014, the share of natural resource rent accounted for 12% of GDP and contributed to 60% of government expenditures (IMF, 2014). Farzanegan (2013) argues that since the 1960s, the government has been using natural resource rent to influence its relationship with its population, by distributing the wealth derived from its natural resources in the shape of government subsidies, public employment and oppressing political antagonists. Thus, natural resource rent has had a significant influence on the economic, social and political landscape of Iran. Fig. 2 shows the breakdown of Iran's merchandise exports, where the fuels and mining products accounted for 73% of total exports with little contribution from manufacturing and agriculture sectors. Smith (2011) introduces a better measure for resource dependence called ‘rent leverage’ that compares the share of per capita oil revenues in GDP per capita terms. In the case of Iran, the rent leverage has been recorded to be more than 30%on average between 1965– 2011.3 In addition, Figs. 3 and 4 graphically represent the composition of Iran's GDP both by the sector of origin and end use. Furthermore, in both metrics the total output represents a significant share of the mining and fuels sectors.

Fuels and mining products

2 3

Middle East and North Africa (MENA) Source: World Bank (2014). For details see Farzanegan (2013).

Manufactures

Fig. 2. Composition of merchandise exports. Source: WTO (2014).

9.1%

50.2% 40.7%

Agriculture

Industry

Services

Fig. 3. GDP composition by sector of origin. Source: World Bank (2014).

-12.7%

20.8%

45.5%

1.2% 31.1% 14.1%

Household consumption

Government expenditures

Investment in fixed capital

Investment in inventories

Exports of goods and services

imports of good and services

Fig. 4. GDP composition by end use. Source: World Bank (2014). 1

Agriculture

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

Is Iran a resource cursed country? The key motivation for this study is to determine if resource abundance crowds-out certain growth friendly factors in the case of Iran. There are various studies on this topic, but they mainly focus on cross-sectional data and have significantly ignored the single country analysis. In contrast, the different economic structures and economic policies have had varied effects on the validation of the resource curse hypothesis at the country and even state level (Lakshmanan and Button, 2009; James and Aadland, 2011). Nevertheless, Iran is an economy which is significantly controlled by state owned firms and their concentration is relatively high in the financial and manufacturing sectors. Iran is also considered a relatively closed economy in terms of trade. On the other hand, the existing literature on the resource curse hypothesis maintains two main streams, the crowding-out and the institutional effects. Here, crowding-out refers to when heavy resource dependency crowds out certain factors which contribute to the economic growth of a country. On the other hand, the institutional effect emphasizes that the impact of resource abundance is subject to the existing institutional quality in an economy. For example, if the institutions are well functioning, then the natural resource abundance positively affects a country's economic growth. Sachs and Warner (2001) explained the crowding-out effect in the ‘Dutch disease’ model and concluded that resource-rich economies plunge into high-price-economies, which further leads to the crowding out of export-led growth in the manufacturing sectors. There are various political and economic factors discussed in earlier literature that influence a country's economic growth (e.g. Olson, 1996; Henisz, 2000; Spero and Hart, 2009) and similar evidence is found in resource-rich developing economies (Shambayati, 1994; Heo and Tan, 2001; Henry and Springborg, 2010; Mahdavi, 2014). Ross (2001) conducts an interesting study that tests the questions: ‘does oil impedes democracy?’ Using timeseries pooled cross-section data of 113 countries, including the Middle-East oil rich economies, the empirical results found a strong link between oil exports and authoritarian rule. Later on, Atkinson and Hamilton (2003) made a claim based on crosscountry regression analysis, suggesting that the so-called resource curse hypothesis is a manifestation of an inability to manage the resource revenues and lagged policy paradigm between macroeconomic policies and resource revenues in a sustainable manner. While attempting to investigate the direct and indirect effect of resource abundance and economic growth, Papyrakis and Gerlagh (2004) found that resource abundance negatively impacts economic growth if taken in isolation and keeping other variables constant, e.g. trade openness, corruption, schooling, and financial development. However, the recent empirical study of Satti et al. (2014) conclude that resource abundance impedes economic growth in Venezuela, even after controller for other determinants of economic growth, e.g. trade openness, capital stock and financial development. Nonetheless, it is further argued that the country specific study is an appropriate tool for impartial conclusions and policy recommendations (Satti et al., 2014). Iran is an economy which is in dire need of research based policy recommendations. This study is also an attempt to not only fill the gap in the resource-abundance and economic growth nexus, but also to assist various national and international policy making bodies that are directly or indirectly linked with Iranian economic development. As such, this study is an empirical investigation to determine both the long- and shortrun associations between natural resource abundance and economic growth by incorporating exports, capital and labor in a production function for the Iranian economy. This study employs the combined cointegration technique developed by Bayer and Hanck (2013). The VECM Granger causality test is also implemented to examine the direction of causality between natural

215

resource abundance and economic growth. This paper is orlganized as follows: Section 2 provides a brief literature review and Section 3 specifies the model and data collection. Section 4 interprets and dicscusses the results. Finally, in Section 5 concludes and suggests policy implications.

2. Review of relevant literature The pioneering study of Sachs and Warner (1995) suggests the surprising feature of economic growth in resource-scarce and resource-rich economies. While analysing a sample of 97 developing economies, their study concludes that the countries with limited natural resources attained higher economic growth than the resource-rich countries. Ross (1999) explains that the resource curse phenomenon in view of political economy and reveals that natural resource abundant countries often fail to manage the link between resource and non-resource sectors, which leads to the ‘Dutch disease’. However, Mikesell (1997) found evidence against the resource-curse via an indirect link between natural resource abundance and economic growth for mineral exporting countries. He goes on to argue that it is not natural resource abundance that directly limits a country's ability to achieve higher economic growth, but its income derived form exports that affects price levels in traded and non-traded sectors, subsequently followed by a growth boom and then long stagnation. At the same time, a series of literature examines the resource curse hypothesis with respect to economic growth, mostly by using cross-country analysis (Frankel, 2010, 2012). The researchers agree to a point that resource abundance does have a non-linear impact on economic growth, but hold little consensus over factors that drive this effect (Doucouliagos and Paldam, 2009; Haber and Menaldo, 2011; Weber, 2012; Cavalcanti et al., 2014). The current thesis on resource curse hypothesis defines both the direct and indirect links between natural resource abundance and economic growth. The empirical evidence is mixed and varies from country to country. For example, the studies of Sachs and Warner (1995, 2001), Papyrakis and Gerlagh (2004), Humphreys et al. (2007) and Mikesell (1997), Stevens (2003), Lederman and Maloney (2007) found strong, as well as weak empirical evidence, of a negative correlation between resource abundance and economic growth, respectively. As such, this contradictory evidence ensures that the topic remains interesting to the academic community while also remaining a mystery for policy makers. Nevertheless, the panel data studies are mostly on the same area and confirm the natural resource curse hypothesis. For example, the recent work of Van der Ploeg (2011) presents a broad survey on the resource-curse hypothesis and finds that the literature suggests that the negative correlation between natural resource abundance and economic growth, are mainly based on the crowding-out4 and institutional effect5 theories. The former concept interprets the idea of a model for the “Dutch disease” explaining that natural resource abundance crowds-out tradable manufactured goods through the additional wealth generated by the sale of natural resources, which in-turn increases country's real exchange rate (Corden and Neary, 1982; Corden, 1984). However, the latter concept focuses on the institutional degradation caused by natural resources in an economy, and argues whether the resources are a blessing or a curse, is determined by the quality of institutions and rejects the Dutch disease model (Mehlum et al., 2006). The Dutch disease hypothesis opines that the economy 4 5

For the crowding-out explanation, refer Sachs and Warner (1995, 2001). For the Institution explanation, refer Mehlum et al. (2006).

216

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

flourishes due to active market participation, called learning by experience, i.e. the manufacturing sector. However, natural resource abundance negatively influences various economic growth factors. For example, Gylfason (2001), Stijns (2006) and Papyrakis and Gerlagh (2004) found that natural resource abundance discourages education, human capital investment and knowledge creation, respectively6. The natural resource windfall anticipates the rise in future income and it reduces savings. The investment to savings adjustment slows down the economic growth process (Papyrakis and Gerlagh, 2006). As a result, natural resource abundant economies usually fail to gain an advantage from technology spillover. The industrial sector, which relies heavily on technological change, is pushed backward along with the exports built up through exchange rate appreciation (Gillis et al., 1996; Gylfason, 2001; Sachs and Warner, 1995, 1999). Papyrakis and Gerlagh (2004) conducted an empirical investigation to check the economic impact of natural resource abundance using crosscountry data. Their results found that natural resource abundance negatively influences the economic development path through various transmission channels and confirmed the existence of the resource curse hypothesis. They also argued that the inclusion of other socio-economic explanatory variables, i.e. trade openness, investment, schooling, and corruption, depict how natural resources have a direct impact on economic growth. Krueger (1990) argued that many of the East and South-East Asia economies have achieved tremendous economic growth in the absence of natural resources, i.e. Japan, Singapore, Taiwan (China), South Korea, Hong Kong S.A.R. (China). These economies successfully harvest the benefits of open economic policies through trade liberalization and export-led-growth policy strategies. The similar types of studies are conducted for various countries and found the natural resource hypothesis to be valid, i.e. Papyrakis and Gerlagh (2007) study the U.S. states, Akinlo (2012) studies the Nigerian oil case. Beck (2011) studies the link between natural resource abundance and financial development. His empirical results conclude that natural resource abundance limits financial development in developing economies. He further explains that due to a weak financial system, capital allocation becomes inefficient and leads to slower economic growth. The recent study of Zuo and Schieffer (2013) empirically examines the resource curse phenomenon at the state level using data from the U.S. Out of the two widely accepted explanations – the crowding-out and institutional effects, their investigation found the crowding-out effect to be dominant over the institution effect in the most U.S states. Furthermore, R&D and investment were found to be the key crowding out factors. A contradicting explanation, disregards the Dutch disease hypothesis and considers the institutional element a key indicator of whether natural resource abundance is a curse or a blessing (Ross, 1999; Atkinson and Hamilton, 2003; Papyrakis and Gerlagh, 2004). Robinson et al. (2006) investigate the political foundation of the resource curse hypothesis based on the link between resource endowment and political incentives in developing economies. They note that politicians over extract natural resources to influence elections and gain power, which further causes the country to misallocate resources. In addition, this practice is accompanied with over-discounting the future and results in inefficiency. Their findings support the institutional explanation and validate the empirical findings of Papyrakis and Gerlagh (2004), Mehlum et al. (2006), and Ross (1999). The main argument made to explain the institutional factor as decisive in determineing whether natural resource abundance is a blessing or curse, was by exemplifying 6 Dutch disease creates crowding effect, and it further leads to adverse effects on indirect transmission channels (i.e. education, human capital investment and knowledge creation) and hinder economic growth (Gylfason, 2001; Stijns, 2006; Papyrakis and Gerlagh, 2004).

resource rich countries, such as Canada, Australia, United States, Denmark and Norway, and there ability to avoid the “Dutch Disease” by building strong institutional systems while they weregoing through economic growth stemming from natural resources (Boschini et al., 2007). However, the countries with weak institutional structure attracts corruption, money laundering, tax evasion, public protest, financial inefficiency and a worsening lawand-order situation, thereby obtaining less gain from a rich endowmenet of natural resources (Acemoglu et al., 2003; Satti et al., 2014). There are two categories of natural resources, point resources and diffuse resources. Point-resources, such as oil and other minerals, generate concentrated and sound revenue patterns that are easily exploitable by small groups of people and have large negative consequences to institutional development (Jensen and Wantchekon, 2004; Mavrotas et al., 2011). Specifically, point-resources create a rentier economy, a weakened political system and distort the distribution of income (Murshed, 2004). However, diffuse-resources, such as agriculture, are better for institutional development, i.e. political. Political economists have defined natural resource abundance in terms of political regimes and governance. For example, Tompson (2005) studied the Russian case and found strong causal linkages between natural resource abundance and poor governance. Similarly, Shao and Qi (2009) confirmed the presence of the resource curse hypothesis, which is harmful for good governance. Shaw (2013) reported that the resource curse hypothesis is present in Azerbaijan due to democracy and bad governance. Satti et al. (2014) validated the occurrence of the resource curse hypothesis i.e. natural recourse abundance impedes economic growth, but financial development and trade openness are contributory factors for domestic production and hence economic growth in Venezuela. There is still a growing amount of literature on the debate of the resource curse hypothesis. It remains unclear whether it is the institutional or Dutch disease hypothesis that causes the resource curse. However, there is no single study conducted on the highly resource rich economy of Iran. Iran is an important country that is facing an extremely slow growth rate for decades, despite having an abundance of natural resources. This study fills this gap in the literature and provides policy recommendations for Iran.

3. Theoretical background and estimation strategy The background and literature discussed in the above section establishes the necessary linkages between natural resource abundance and economic growth. The discussion further develops a valid research rationale of why it is imperative to test the resource curse hypothesis in Iran. In this context, the prime objective of the this study is to investigate the relationship between natural resource abundance and economic growth by incorporating exports, capital and labor as additional explanatory variables to test the resource curse hypothesis in Iran. This study uses updated yearly time-series data over the extended period of 1965–2011. The length and frequency of the data is based on availability.7 The World Development Indicators (official CD-ROM, 2014) are combed to collect the time-series data on natural resource abundance, real GDP (local currency), real exports (local currency), real capital (local currency) and labor. We have used population data to 7 Our study empirically establishes resource course hypothesis using a long dataset. The reason for the choice of choosing larger sample size is that the use of long dataset not only increases the total number of observation but also enables the empirical estimation to have higher degrees of freedom. To some extent, it reduces noise coming from the individual time series cointegrated regressions and also establishes the long-run relationships between the series.

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

transform all of the variables into per capita terms. The functional form of the model is given as follows:

Yt = f (NRt , Et , Kt , L t )

(1)

ln Yt = β1 + β2 ln NR t + β3 lnEt + β4 ln Kt + β5 lnL t + μi

(2)

We transformed all of the variables into a log-linear specification.8 ln Yt is the natural log of real GDP per capita and is a proxy for economic growth, lnNRt is the natural log of natural resource abundance (per capita), lnEt is the natural log of real exports per capita, ln Kt is the natural log of real capital use and labor use and is measured by labor per capita ( lnLt ) while μi is the error term. β1 is an autonomous coefficient or constant term and β2, β3, β4 and β5 are induced coefficients for measuring the relative influences of natural resource abundance, real exports, real capital and labor use in the aggregate production function shown in Eq. 2. If β2 < 0 , it indicates that natural resource abundance has an adverse effect on real economic growth in Iran. If β3 < 0, it shows the negative contribution of real exports on real economic growth, indicating that real exports are contributing less to Iran’s real economic growth. If β4 and β5 < 0, this indicates that both capital and labor per capita are not being used effectively in the production process and thereby have adverse effects on economic growth. Given this hypothetical setting, it is again important to highlight the relative strength of real GDP that is influenced by the rest of the independent variables, i.e. real natural resource abundance per capita, real exports per capita, real capital use per capita and labor use. We believe that an economy with a stock of natural resources, will see these resources contribute to higher economic growth in the transition phase, provided that the economy is making efficient use of its endowment of natural resources for its production and investment activities. As far as real exports are concerned, it is expected that Iran will receive more export revenue from better economic production induced by a larger endowment of natural resources, which will also lead to higher economic growth. We also believe that capital and labor combine together to provide higher economic growth if they are used efficiently in the process of aggregate economic activity. 3.1. Bayer–Hanck combined cointegration modelling The applied econometric theory claims that the linear combination of a series has a lower order of integration if the time-series are integrated at I(1) or (2). Engle and Granger (1987) pioneered the cointegration approach to examine the long-run relationship between series. This test of cointegration requires that all series must have a unique order of integration. The Engle–Granger cointegration approach is suitable when the data sets have limited length as most economic time-series are found to be analyzed in a partial macroeconomic framework. However, the problem with the Engle–Granger cointegration approach is that it provides biased empirical results due its low explanatory power properties. In the early 1990s, Johansen (1991) introduced a new test of cointegration titled “Johansen maximum Eigen-value test”. This test of cointegration is more preferable to researchers because it permits more than one cointegrating relationship between the series. Phillips and Ouliaris (1990) developed another approach for investigation of cointegration between series, which is termed the Phillips–Ouliaris cointegration test. The Error Correction Model (ECM) based F-test was developed by Boswijk (1994), and the ECM based t-test by Banerjee et al. (1998). Bayer–Hanck (2013) combined cointegration approach employs 8

We use natural logarithm for all level series considered in the estimation of resource curse hypothesis for the sake of producing smoothness in the time series data.

217

different tests that suggest different conclusions. The null of the no-cointegration of Bayer–Hanck combined cointegration test is purely based on Engle and Granger (EG, 1987), Johansen (JOH, 1991), Boswijk (BO, 1994), and Banerjee et al. tests (BDM, 1998). The Bayer–Hanck test jointly determines the test-statistics of Engle and Granger, Johansen, Boswijk, and Banerjee et al. tests. This cointegration approach combines the empirical results of various individual cointegration tests to obtain comprehensive cointegration findings. We apply this approach of combined cointegration to determine whether cointegration is present between natural resource abundance and economic growth in the case of Iran. The Bayer and Hanck (2013) combined cointegration technique is computed based on Fisher's (1932) formulas to obtain the statistical significance level i.e. p-values of the single cointegration test, and the formula is shown below:

EG − JOH = − 2 ⎡⎣ ln (pEG ) + (pJOH ) ⎤⎦

(3)

EG − JOH − BO − BDM = − 2 ⎡⎣ ln (pEG ) + (pJOH ) + (pBO ) + (pBDM ) ⎤⎦

(4)

The probability values of different individual cointegration tests include EG (1987), JOH (1991), BO (1994), and BDM (1998) are followed by pEG , pJOH , pBO and pBDM . This indicates that the p-values of various individual cointegration tests are involved in the Fisher (1932) equations. As far as the combined cointegration testing approach is concerned in this analysis, we believe the null hypothesis of no cointegration is rejected if the estimated Fisher statistics exceed the critical values provided by the Bayer and Hanck (2013) estimation.

4. Results and discussion The results reported in Table 1 show that an abundance of natural resources and exports are highly volatile compared to capital use. Labor shows less variation compared to economic growth. Jarque-Bera estimates confirm the normality of the variables, i.e. economic growth, natural resource abundance, exports, capital use and labor. The correlation analysis indicates that natural resource abundance and economic growth are negatively correlated. The correlation between exports (capital use) with economic growth is positive. Labor force is positively correlated with economic growth. A positive correlation is found in exports, capital use and labor with natural resource abundance. Capital and labor are positively related to exports. Labor and capital are also positively correlated. The next step is to test the integrating properties of the Table 1 Descriptive statistics and correlation matrix. Variables

ln Yt

ln NRt

ln Et

ln Kt

ln Lt

Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability ln Yt ln NRt ln Et ln Kt ln Lt

15.4664 15.4218 15.8075 15.0898 0.1922 0.1161 2.0171 1.7849 0.4096 1.0000
0.5159 0.5650 0.7535 0.2932

18.6489 18.6486 19.6682 16.6028 0.6728 -0.7194 3.6598 4.3847 0.1116

13.8995 13.9681 14.9423 12.0034 0.6796 -0.6113 2.9682 2.6178 0.2701

14.3452 14.3154 15.1591 13.7162 0.3460 0.2008 2.4199 0.8711 0.6468

4.0422 3.9831 4.2780 3.9270 0.1206 0.9566 2.3084 3.2434 0.2673

1.0000 0.4035 0.2676 0.1700

1.0000 0.3037 0.1832

1.0000 0.2207

1.0000

218

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

Table 2 Unit root analysis.

Table 4 The Results of Bayer and Hanck cointegration Analysis.

Variable

MZa

ln Yt ln NRt ln Et ln Kt ln Lt Δ ln Yt Δ ln NRt Δ ln Et Δ ln Kt Δ ln Lt


10.3295
6.85153
5.13976
12.0830
13.0502
27.7575
19.8069
23.4897
23.7808
42.6537

(1) (2) (3) (1) (4) (3)n (2)nn (2)nn (2)nn (4)n

MZt

MSB

MPT

Estimated models

EG–JOH

EG–JOH–BO–BDM

Lag order

Cointegration


2.2288
1.8416
1.5742
2.4326
2.4996
3.7254
3.1364
3.4268
3.4476
4.5773

0.2157 0.2687 0.3063 0.2013 0.1915 0.1342 0.1583 0.1458 0.1449 0.1073

9.02569 13.3085 17.6024 7.6771 7.2871 3.2829 4.6641 3.8807 3.8353 2.3475

Yt = f (NRt , Et , Kt , Lt )

19.653a

130.177a

3

Exists

NRt = f (Yt , Et , Kt , Lt )

17.193a

33.063a

3

Exists

Et = f (Yt , NRt , Kt , Lt )

19.980a

41.474a

3

Exists

Kt = f (Yt , NRt , Et , Lt )

17.407a

79.427a

3

Exists

Lt = f (Yt , NRt , Et , Kt )

13.478

18.499

4

Not exist

Table 5 Long run analysis.

Note: Lag order is shown in parenthesis. n

Represent significant at 1% levels respectively. Represent significant at 5% levels respectively.

nn

Dependent variable ¼ ln Yt

variables by applying the N–P unit root test developed by Ng and Perron (2001). This test is suitable and provides efficient empirical results for the small sample in our data set. The NP unit root test is superior to the ADF, PP and DF-GLS unit root test due to its explanatory power. The results are reported in Table 2, which indicate that economic growth, natural resource abundance, exports, capital and labor contain an a unit root problem at level with the intercept and trend. However, it is noted that all of the variables are found to be stationary at the 1st difference. This shows that economic growth, natural resource abundance, exports, capital and labor have a unique order of integration i.e. I(1). The unique order of integration leads us to apply the combined cointegration approach developed by Bayer and Hanck (2013). An appropriate lag length is necessary to apply the Bayer– Hanck (2013) combined cointegration approach to examine whether cointegration is present amongst the variables. We have chosen Akaike information criterion to select an appropriate lag length. The results of the Bayer–Hanck (2013) combined cointegration approach are sensitive to lag order selection (Satti et al., 2014). The results reported in Table 3 show that lag 3 should be used while conducting the empirical analysis. The results of the Bayer–Hanck (2013) combined cointegration approach are reported in Table 4. The results reveal that we may reject the hypothesis of no-cointegration as economic growth, natural resource abundance, exports and capital are used as dependent variables. Our calculated Fisher statistics for EG–JOH and EG–JOH–BO–BDM exceeds the critical values at the 1% level. We failed to reject the null hypothesis of no cointegration as economic growth, natural resource abundance, exports and capital are used as independent variables. This shows the presence of 4 cointegrating vectors Table 3 Lag length selection. VAR lag order selection criteria Lag

LR

FPE

AIC

SC

HQ

0 1 2 3

NA 284.8901 137.2961 77.24579a

2.81e
08 1.83e
11 5.23e
13 7.83e
14a


3.1967
10.5477
14.1690
16.2455a


2.9834
9.2680
11.8230
12.8331a


3.1201
10.0885
13.3273
15.0211a

LR: sequential modified LR test statistic (each test at 5% level) FPE: Final prediction error AIC: Akaike information criterion SC: Schwarz information criterion HQ: Hannan-Quinn information criterion a

Indicates lag order selected by the criterion.

a Represents significant at 1% level. Critical values at 1% level are 15.701 (EG– JOH) and 29.85 (EG–JOH–BO–BDM) respectively. Lag length is based on minimum value of AIC.

Variables

Coefficient

Std. Error

t-Statistic

Prob. value

Constant ln NRt ln E ln Kt ln Lt

8.6920n
0.4662nn 0.1463n 0.3210n 0.2485n

0.3925 0.2228 0.0301 0.0303 0.0749

22.1447
2.0922 4.8532 10.5757 3.3165

0.0000 0.0435 0.0000 0.0000 0.0021

Diagnostic tests R2 F-statistic

0.9594 213.0556n

Test

χ 2 NORMAL

F-statistic 1.5785

Prob. value (0.4541)

χ 2 SERIAL

2.1163

(0.0898)

χ 2 ARCH

0.0366

(0.8491)

χ 2 HETERO

5.2349

(0.2640)

χ 2 RAMSEY

0.6911

(0.4114)

nnn

Denote the significant at 10% levels respectively.

χ 2 NORM is for normality test, χ 2 SERIAL for LM serial correlation test, χ 2 ARCH for autoregressive conditional heteroskedasticity,

χ 2 HETERO for white hetero-

skedasticity and χ 2 REMSAY for Resay Reset test. n

Denote the significant at 1% levels respectively. Denote the significant at 5% levels respectively.

nn

which validate the existence of cointegration between the variables for time period of 1965–2011 in the case of Iran (Tables 5 and 6). In the long run, we find that the natural resource abundance negatively affects economic growth. This notion validates the resource curse hypothesis in Iran. A 1% increase in natural resource abundance impedes economic growth by 0.47%, ceteris paribus. This result further implies that the Iranian economy is suffering from the Dutch disease. The large extraction of natural resources, i.e. Oil and natural gas, leads to large inflows of foreign currency that ultimately results in an unnecessary appreciation of the home currency. Hence, the country looses price competitiveness in other export products and adversely affects GDP. In 2010, the oil and gas sector represented 74% of total exports and 50% of Government expenditures in Iran. In 2013, Iran’s total natural resource rent stood at 39.4% of GDP. The country is required to focus on manufacturing products other than natural resources. This empirical evidence is consistent with the findings of Papyrakis and Gerlagh (2004); Papyrakis and Gerlagh (2007); Asekunowo and Olaiya (2012) and Satti et al. (2014) in the case of the USA, Nigeria and Venezuela. The relationship between exports and economic growth is positive and statistically significant at the 1% level. Exports stimulate economic growth. Keeping other things same, a 1% increase in economic growth is augmented by a 0.15% increase in exports. This phenomenon endorses our previous statement and

K. Ahmed et al. / Resources Policy 49 (2016) 213–221 20

Table 6 Short run analysis.

15 10

Dependent Variable ¼ Δ ln Yt

5

Variables

Coefficient

Std. Error

t-Statistic

Prob.

Constant Δ ln NRt Δ ln E Δ ln Kt Δ ln Lt ECMt − 1


0.0017
0.0121 0.0741nn 0.24718n 0.3755
0.6094n

0.0066 0.0159 0.0281 0.0238 0.3915 0.1164


0.2685
0.0007 2.6299 10.3659 0.9591
5.2358

0.7899 0.9937 0.0129 0.0000 0.3445 0.0000

Diagnostic Tests R2 F-statistic Test

219

0 -5 -10 -15 -20 1980

1985

1990 CUSUM

1995

2000

2005

2010

2005

2010

5%Significance

Fig. 5. Cumulative sum of recursive residual. 0.7747 19.9524n

1.4

χ 2 NORMAL

F-statistic 1.4336

Prob.value (0.4858)

χ 2 SERIAL

1.7838

(0.1484)

0.8

χ 2 ARCH

2.0038

(0.1331)

0.6

χ 2 HETERO

1.3504

(0.2388)

0.4

χ 2 RAMSEY

0.1372

(0.7135)

0.2

1.2 1.0

0.0 nnn

Denote the significant at 10% levels respectively.

χ 2 NORM is for normality test, χ 2 SERIAL for LM serial correlation test, χ 2 ARCH for autoregressive conditional heteroskedasticity,

χ 2 HETERO for white hetero-

skedasticity and χ 2 REMSAY for Resay Reset test. n

Denote the significant at 1% levels respectively. Denote the significant at 5% levels respectively.

-0.2 -0.4 1980

1985

1990

1995

CUSUM of Squares

2000 5% Significance

Fig. 6. Cumulative sum of square of recursive residual.

nn

reinforce the argument that non-oil and -gas sectors need to be mobilized to have an increasing share of the country’s total exports. The most important thing in achieving such targets, is to increase the participation of the private sector. This finding is consistent with Shahbaz (2012), who noted that exports play a significant role in enhancing domestic output and hence economic growth for Pakistan's economy. Capital plays a dominant role in increasing domestic production and results in increased economic growth. A 1% increase in capital use is linked with economic growth by 0.32%. Labor is positively and significantly associated with economic growth. A 0.25% increase in economic growth is augmented by a 1% spur in the labor force. In the short run, the effect of natural resource abundance on economic growth is negative, but statistically insignificant. The relationship between exports and economic growth is positive and significant at the 5% level. Capital is positively and significantly associated with economic growth. Labor has positive, but an insignificant effect on economic growth. The long-run coefficients are greater than short-run parameters, which confirm the reliability and stability. The estimate of the ECMt
1 is
0.6094 and significant at the 1% level. This corroborates our long-run relationship between natural resource abundance and economic growth in Iran. We conclude that following the economic growth function, short-run deviations are corrected by 60.94% towards a path for long-run equilibrium. While checking the robustness of the model through various statistical diagnostics tests and statistics, we find that the error term is normally distributed and the notion is confirmed by the Jarque–Bera test and there is no evidence for white heteroskedasticity. The short-run analysis is suffering from the serial correlation problem and the ARCH problem. However, the Ramsey reset test confirms that the specification of the model is well designed. The CUSUM and CUSUMsq test are applied for the stability of the short-run and long-run parameters. The results are reported in Figs. 2 and 3. We find that the graphs of the CUSUM and CUSUMsq test lie within the 5% critical bounds. This confirms the

stability of the parameters (Figs. 5 and 6). The short-run, as well as long-run causal relationship between natural resource abundance and economic growth, is investigated by applying the VECM Granger causality test. The results are reported in Table 7 and suggest a feedback effect between natural resource abundance and economic growth in the long-run path, i.e. natural resource abundance causes (declines) economic growth and in return, economic growth causes (declines) natural resource abundance. The relationship between exports and economic growth is bidirectional. Capital use causes economic growth and economic growth causes capital use in the Granger sense. The unidirectional causality is found running from labor to economic growth, natural resource abundance, exports and capital use. The feedback effect is validated between exports and natural resource abundance. The causality between capital use and natural resource abundance is bidirectional. In the short-run, exports cause economic growth and economic growth causes exports in the Granger sense. The feedback effect exists between capital use and economic growth. Exports lead to natural resource abundance and the same is true of the opposite side, i.e. bidirectional causality. The unidirectional causality exists running from labor to natural resource abundance. A summary of the causal relationship is provided in Table 8.

5. Conclusion and policy implications This paper empirically examines the statistical relationship between natural resource abundance and economic growth under the resource curse hypothesis in the case of Iran. Using updated time-series data of annual frequency over the extended period of 1965–2011, the study adopts the Ng–Perron unit root test (Ng and Perron, 2001) followed by the Bayer–Hanck combined cointegration approach (Bayer and Hanck, 2013) to test the stationary property and long-run relationship between underlying variables, respectively. Having confirmed cointegration among the variables, the long-run and short-run analysis is applied to test the resource curse hypothesis. Finally, the VECM Granger causality test is used

220

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

Table 7 The VECM Granger causality analysis. Dependent Variable

Direction of causality Short run

Δ ln Yt − 1

Long run

Δ ln NRt − 1

Δ ln Lt − 1

ECTt − 1

nn

nn

Δ ln Et − 1

Δ ln Kt − 1

Δ ln Yt

–

0.0013 [0.9986]

3.4030 [0.0470]

3.5858 [0.0391]

0.3037 [0.7404]


0.5389n [
4.5587]

Δ ln NRt

0.2561 [0.7758]

–

3.8596nn [0.0356]

0.3605 [0.7005]

5.1959nn [0.0120]


0.6934n [
3.1720]

Δ ln Et

5.9846n [0.0069]

17.0707n [0.0000]

–

1.0923 [0.3493]

1.0788 [0.3537]


0.9523n [
4.12087]

Δ ln Kt

3.3231nn [0.0497]

0.0272 [0.9732]

1.9088 [0.1670]

–

0.4199 [0.6611]


0.7010n [
3.5471]

Δ ln Lt

0.0312 [0.9692]

0.0008 [0.9992]

0.3602 [0.7006]

0.0318 [0.9687]

–

–

nnn

Shows significance at 10% levels respectively. n

Shows significance at 1% levels respectively. Shows significance at 5% levels respectively.

nn

Table 8 Summary (The VECM Granger causality test). Direction of causality

ln ln ln ln ln ln ln ln ln ln ln ln ln ln ln ln ln ln ln ln

Yt Granger causes ln NRt Yt Granger causes ln Et Yt Granger causes ln Kt Yt Granger causes ln Lt NRt Granger causes ln Yt NRt Granger causes ln Et NRt Granger causes ln Kt NRt Granger causes ln Lt Et Granger causes ln Yt Et Granger causes ln NRt Et Granger causes ln Kt Et Granger causes ln Lt Kt Granger causes ln Yt Kt Granger causes ln NRt Kt Granger causes ln Et Kt Granger causes ln Lt Lt Granger causes ln Yt Lt Granger causes ln NRt Lt Granger causes ln Et Lt Granger causes ln Kt

Short run (F-statistics) No At 1% At 5% No No At 1% No No At 1% At 5% At 5% No At 1% No No No No At 5% No No

Long tun ( ECTt − 1)

At 1% significance level At 1% significance level At 1% No At 1% significance level At 1% At 1% No significance level At 1% significance level At 1% significance level At 1% No significance level At 5% At 5% At 5% No At 5% significance level At 5% At 5% At 5%

significance level significance level significance level significance level significance level significance level significance level significance level significance level significance level significance level significance level significance significance significance significance

level level level level

to examine the causal links between the variables. The cointegration test results confirm the presence of cointegration between natural resource abundance, economic growth, exports, capital and labor. The long-run analysis results find that natural recourses have an adverse impact on economic growth and validate the resource curse hypothesis in Iran. Exports, capital use and labor have a positive impact on economic growth. The causality analysis indicates a feedback effect between natural resource abundance and economic growth. Exports cause economic growth and in response economic growth causes exports in the Granger sense. The unidirectional causality exists running from labor to economic growth. The bidirectional causality also exists between capital use and economic growth. Resource abundant economies mostly rely on the exports of mineral resources. In such economies, very little effort is made to boost the export potential that exists in other sectors (Sachs and

Warner, 1995). This is also true in the case of Iran. Our findings suggest that Iran's economy relies heavily on natural resource endowment, which hampers its economic growth in the long-run. Alternatively, diversifying the economy can substantially increase economic growth. Investing in competitive and non-hydrocarbon exports can achieve such an objective. Unlike the oil rich economies in the gulf, Iran has the potential to become involved in intraindustry trade to achieve trade specialization. Trade diversification measures anticipate new investments in emerging profitable businesses and produce new employment opportunities. Reduced subsidies and government interventions enable domestic businesses to compete in both national and international markets and improve the gross capital formation in the country. Prudent fiscal policy coupled with a natural resource fund can be used to encourage investment in promising sectors of an economy. The role of capital in boosting economic growth is wellestablished. Efficient capital increases output, thereby economic growth could be improved. Providing subsidies and tax exemption to businesses will help build capital in the economy. New and efficient capital would save natural resources in output production on one hand and would cause a decline in the cost of production on the other. Building capital would also help reduce reliance on natural resource exports through diversifying economy. The study is conducted soon after the end of the decades long economic sanctions on Iran. Hence, the findings possess critical policy implications in the wake of Iran's new role in the world economy.

References Acemoglu, D., Johnson, S., Robinson, J., 2003. Disease and development in historical perspective. J. Eur. Econ. Assoc. 1 (2–3), 397–405. Akinlo, A.E., 2012. How important is oil in Nigeria's economic growth? J. Sustain. Dev. 5 (4), p165. Asekunowo, V.O., Olaiya, S.A., 2012. Crude oil revenue and economic development in Nigeria (1974–2008). OPEC Energy Rev. 36 (2), 138–169. Atkinson, G., Hamilton, K., 2003. Savings, growth and the resource curse hypothesis. World 31 (11), 1793–1807. Banerjee, A., Dolado, J., Mestre, R., 1998. Error-correction mechanism tests for cointegration in a single-equation framework. J. Time Ser. Anal. 19 (3), 267–283. Bayer, C., Hanck, C., 2013. Combining non-cointegration tests. J. Time Ser. Anal. 34 (1), 83–95. Beck, J.S., 2011. Cognitive Behavior Therapy: Basics and Beyond. Guilford Press. Boschini, A.D., Pettersson, J., Roine, J., 2007. Resource curse or not: a question of

K. Ahmed et al. / Resources Policy 49 (2016) 213–221

appropriability. Scand. J. Econ. 109 (3), 593–617. Boswijk, H.P., 1994. Testing for an unstable root in conditional and structural error correction models. Journal. Econom. 63 (1), 37–60. Brunnschweiler, C.N., Bulte, E.H., 2008. The resource curse revisited and revised: a tale of paradoxes and red herrings. J. Environ. Econ. Manag. 55 (3), 248–264. Cavalcanti, D.V., Tiago, V., Mohaddes, K., Raissi, M., 2014. Commodity price volatility and the sources of growth. J. Appl. Econom. http://dx.doi.org/10.1002/jae.2407 Collier, P., Goderis, B., 2008. Commodity prices, growth, and the natural resource curse: reconciling a conundrum. Growth, and the Natural Resource Curse: Reconciling a Conundrum (June 5, 2008). Corden, W.M., 1984. Booming sector and Dutch disease economics: survey and consolidation. Oxf. Econ. Pap., 359–380. Corden, W.M., Neary, J.P., 1982. Booming sector and de-industrialisation in a small open economy. Econ. J., 825–848. Doucouliagos, H., Paldam, M., 2009. The aid effectiveness literature: The sad results of 40 years of research. J. Econ. Surv. 23 (3), 433–461. Dubé, J., Polèse, M., 2015. Resource curse and regional development: does Dutch disease apply to local economies? Evidence from Canada. Growth Change 46 (1), 38–57. Energy International Agency (EIA), 2014. 〈https://www.eia.gov/beta/international/ country.cfm?iso ¼ IRN〉 (accessed: 10.05.16). Engle, R.F., Granger, C.W., 1987. Co-integration and error correction: representation, estimation, and testing. Econ.: J. Econ. Soc., 251–276. Farzanegan, M.R., 2013. Oil and the Future of Iran: a Blessing or a Curse. Legatum Institute Future of Iran series, London. Fisher, Irving, 1932, Booms and depressions: Some first principles (Adelphi, New York).Energy International Agency (EIA), 2014. 〈https://www.eia.gov/beta/in ternational/country.cfm?iso ¼IRN〉 (Accessed: 10.05.2016). Frankel, J.A., 2010. The Natural Resource Curse: A Survey (No. w15836). National Bureau of Economic Research. Frankel, J.A., 2012. The Natural Resource Curse: A Survey of Diagnoses and Some Prescriptions. Gillis, K.D., Mößner, R., Neher, E., 1996. Protein kinase C enhances exocytosis from chromaffin cells by increasing the size of the readily releasable pool of secretary granules. Neuron 16 (6), 1209–1220. Gylfason, T., 2001. Natural resources, education, and economic development. Eur. Econ. Rev. 45 (4), 847–859. Haber, S., Menaldo, V., 2011. Do natural resources fuel authoritarianism? A reappraisal of the resource curse. Am. Polit. Sci. Rev. 105 (01), 1–26. Henisz, W.J., 2000. The institutional environment for economic growth. Econ. Polit. 12 (1), 1–31. Henry, C.M., Springborg, R., 2010. Globalization and the Politics of Development in the Middle East vol. 1. Cambridge University Press. Heo, U., Tan, A.C., 2001. Democracy and economic growth: a causal analysis. Comp. Polit., 463–473. Humphreys, M., Sachs, J., Stiglitz, J.E. (Eds.), 2007. Escaping the Resource Curse. Columbia University Press, New York, pp. 11–13. International Monetary Fund (IMF), 2014. 〈https://www.imf.org/external/pubs/ft/ scr/2014/cr1493.pdf〉 (accessed: 10.05.16). James, A., 2015. The resource curse: a statistical mirage? J. Dev. Econ. 114, 55–63. James, A., Aadland, D., 2011. The curse of natural resources: an empirical investigation of U.S. counties. Resour. Energy Econ. 33 (2), 440–453. Jensen, N., Wantchekon, L., 2004. Resource wealth and political regimes in Africa. Comp. Polit. Stud. 37 (7), 816–841. Johansen, S., 1991. Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models. Econ.: J. Econ. Soc., 1551–1580. Krueger, A.O., 1990. Government Failures in Development (No. w3340). National Bureau of Economic Research. Lakshmanan, T.R., Button, K.J., 2009. Institutions and regional development. In: Capello, R., Nijkamp, P. (Eds.), Handbook of Regional Growth and Development Theories. Edward Elgar, Cheltenham, UK and Northampton, MA, USA, pp. 443–460. Lederman, D., Maloney, W.F. (Eds.), 2007. Natural Resources, Neither Curse nor Destiny. World Bank Publications. Mahdavi, P., 2014. Why do leaders nationalize the oil industry? The politics of resource expropriation. Energy Policy 75, 228–243. Mavrotas, G., Murshed, S.M., Torres, S., 2011. Natural resource dependence and economic performance in the 1970–2000 period. Rev. Dev. Econ. 15 (1), 124–138.

221

Mehlum, H., Moene, K., Torvik, R., 2006. Institutions and the resource curse. Econ. J. 116 (508), 1–20. Mikesell, R.F., 1997. Explaining the resource curse, with special reference to mineral-exporting countries. Resour. Policy 23 (4), 191–199. Murshed, S.M., 2004. When does natural resource abundance lead to a resource curse? (No. 24137). International Institute for Environment and Development, Environmental Economics Programme. Ng, S., Perron, P., 2001. Lag length selection and the construction of unit root tests with good size and power. Econometrica, 1519–1554. Olson, M., 1996. Distinguished lecture on economics in government: big bills left on the sidewalk: why some nations are rich, and others poor. J. Econ. Perspect., 3–24. Papyrakis, E., Gerlagh, R., 2004. The resource curse hypothesis and its transmission channels. J. Comp. Econ. 32 (1), 181–193. Papyrakis, E., Gerlagh, R., 2006. Resource windfalls, investment, and long-term income. Resour. Policy 31 (2), 117–128. Papyrakis, E., Gerlagh, R., 2007. Resource abundance and economic growth in the United States. Eur. Econ. Rev. 51 (4), 1011–1039. Phillips, P.C., Ouliaris, S., 1990. Asymptotic properties of residual based tests for cointegration. Econ.: J. Econ. Soc., 165–193. Robinson, J.A., Torvik, R., Verdier, T., 2006. Political foundations of the resource curse. J. Dev. Econ. 79 (2), 447–468. Ross, M.L., 1999. The political economy of the resource curse. World Polit. 51 (02), 297–322. Ross, M.L., 2001. Does oil hinder democracy? World Polit. 53 (03), 325–361. Sachs, J.D., Warner, A.M., 1995. Natural Resource Abundance and Economic Growth (No. w5398). National Bureau of Economic Research. Sachs, J.D., Warner, A.M., 1999. The big push, natural resource booms and growth. J. Dev. Econ. 59 (1), 43–76. Sachs, J.D., Warner, A.M., 2001. The curse of natural resources. Eur. Econ. Rev. 45 (4), 827–838. Satti, S.L., Farooq, A., Loganathan, N., Shahbaz, M., 2014. Empirical evidence on the resource curse hypothesis in oil abundant economy. Econ. Model. 42, 421–429. Shahbaz, M., 2012. Does trade openness affect long run growth? Cointegration, causality and forecast error variance decomposition tests for Pakistan. Econ. Model. 29 (6), 2325–2339. Shambayati, H., 1994. The rentier state, interest groups, and the paradox of autonomy: state and business in Turkey and Iran. Comp. Polit., 307–331. Shao, S., Qi, Z.Y., 2009. A theoretic interpretation and empirical test on R&D behavior of energy-output-oriented cities based on resource curse hypothesis. J. Financ. Econ. 1, 008. Shaw, D.L., 2013. Good governance in the post-Soviet south: testing theories of the resource curse in Azerbaijan. J. Polit. Int. Stud. 9, 520–561. Smith, B., 2011. Looney, R.E. (Ed.), Oil and politics in South East Asia, pp. 206–218. Spero, J.E., Hart, J., 2009. The Politics of International Economic Relations. Cengage Learning. Stevens, P., 2003. Resource impact: curse or blessing? A literature survey. J. Econ. Lit. 9, 3–42. Stijns, J.P., 2006. Natural resource abundance and human capital accumulation. World Dev. 34 (6), 1060–1083. Tompson, W., 2005. The political implications of Russia’s resource-based economy. In: “Europe: Our Common Home”, the Seventh International Council for Central and East European Studies (ICCEES) Congress, 26–30 July 2005, Berlin. Van der Ploeg, F., 2011. Natural resources: Curse or blessing? J. Econ. Lit., 366–420. Van der Ploeg, F., Poelhekke, S., 2009. Volatility and the natural resource curse. Oxf. Econ. Pap. 61 (4), 727–760. Weber, J.G., 2012. The effects of a natural gas boom on employment and income in Colorado, Texas, and Wyoming. Energy Econ. 34 (5), 1580–1588. Williams, A., 2011. Shining a light on the resource curse: an empirical analysis of the relationship between natural resources, transparency, and economic growth. World Dev. 39 (4), 490–505. World Bank, 2014. World Development Indicators. 〈http://data.worldbank.org/sites/ default/files/wdi-2014-book.pdf〉 (accessed: 10.05.16). World Trade Organization (WTO), 2014. 〈http://stat.wto.org/CountryProfile/ WSDBCountryPFView.aspx?Country ¼ IR&Language¼E〉 (accessed: 10.05.16). Zuo, N., Schieffer, J., 2013. Crowding-out Effect or Institutions? The Resource Curse Revisited with an Investigation of US States. In: Southern Agricultural Economics Association Meeting, Orlando, FL, February. pp. 3–5.