The Research Hypothesis and Data

Resource Curse and Economic Growth

Introduction

There are plenty of reasons to explain the unequal economic growth problem among countries and no doubt it is still an important issue even for countries with abundant natural resources. Although developed, developing, or least developed, are in various stages of economic growth and executing reforms in this direction, recent development economics discoveries show that in period from 1960 to 1990 resource poor countries on average grew 2-3 times more than resource abundant countries (Sachs and Warner, 1995; Auty, 2001)[1].

Yet it does not mean that we come to conclusion that all resource rich countries are cursed. There are examples of economic growth with abundant natural resources such as Australia, Canada, Norway, Botswana, Mauritius and Chile. The Figure 1 below describes the growth performance facts of 87 developing and emerging economies. The averages (medians) of real income per capita growth rates by groups from 1991 to 2010 for emerging economies are plotted, to make brief analysis of the hypothesis stated above. The data covers 87 countries and grouped according to their export structures. The classification is identified and suggested by Woolcock et al. (2001) and Isham et al. (2005).

It is observed from the Figure 1 that manufacturing goods exporting countries have the highest average income per capita growth relative to other three groups of countries except in years of economic crisis (e.g. 1998, 2007-2009). It is also documented that point source and coffee/cocoa products concentrated countries are performing the slowest growth. The average of median growth rates during 2000-2010 are 4.10%, 2.95%, 2.33%, 1.81% in manufacturing, diffuse, point source and coffee/cocoa goods concentrated countries respectively. However, the data since 2000 shows that the resource rich economies are catching up manufacturing goods concentrated ones. As illustrated in Figure 1 upward sloping trend is observed in resource rich countries, especially in point source and diffuse resource rich economies. This goes in line with World Bank study, which documented that five out of 82 countries are most resource rich countries and those countries are among the top 15 leading economies according to income per capita level (World Bank, 1994). Has anything changed since 2000 in favor of resource rich countries?

Figure 1: Smoothed Median Growth Rates for 87 Developing and Emerging Economies,

1991-2010

Source: Own Calculation based on IMF World Economic Outlook Database

To explain these disparities of economic growth in different economies grouped according to their export concentration, this study focuses on the extent of variation in economic growth, especially before and since 2000, due to institutional arrangements. As a first take, average yearly GDP per capita growth versus log of Resource Rents (RR) is plotted in Figure 2 and covers 79 countries out of 87 discussed above[2]. The data for Resource Rents (RR) are retrieved from World Bank database[3]. The data has been split into two subsamples: panel (a) the yearly average economic growth from 1991 to 2000 versus log of Resource Rents (as a % of GDP) in corresponding years and panel (b) the yearly average economic growth from 2001 to 2010 versus log of Resource Rents (as a % of GDP) in corresponding years. Now the sign of resource curse is observed in panel (a) only. Indeed, panel (a) gives strong evidence of resource curse, while it is not observed in panel (a). Surprisingly, the weak positive relationship is spotted in panel (b).

Figure 2: GDP per capita growth and Log of Resource Rents % in GDP (1991-2001 versus 2001-2010)


(a) GDP per capita and Log of Resource Rents share in GDP during 1991-2000	(b) GDP per capita and Log of Resource Rents share in GDP during 2001-2010

Figure 2: Growth and Resource Rents share in GDP for the periods (a) 1991-2000 and (b) 2001-2010.

Notes: World Bank data of Resource Rents (% of GDP) and GDP per capita growth.

As a further illustration of how resource rents explain economic growth via institutional arrangements during 2001-2010 in Figure 3, the data has been divided into two groups according to institutional quality level controlling for manufacturing goods concentrated countries[4]. Thus, the data set consists of 73 countries, divided into two groups of approximately equal size in panel (a) 36 countries with good institutions and (b) 37 countries with bad institutions[5]. It is observed from the Figure 3 below that the resource curse is present in countries with bad institutions, while the resource concentrated countries with good institutions demonstrate good economic performance. This is also observed in many empirical and theoretical studies carried out earlier[6]. The major distinction between grabber friendly and producer friendly institutions is important as they impact economic growth in different ways. Grabber friendly institutions are particularly bad for economic growth because rent-seeking associated with bad institutions put limited entrepreneurial resources involved in production into inefficient activities.

Figure 3: GDP per capita growth and Log of Resource Rents % in GDP (Good versus Bad Institutions, 2001-2010)


(a) With good institutions	(b) With bad institutions

Figure 3: Growth versus log of Resource Rents % in GDP (a) with good institutions and (b) bad institutions.

Notes: World Bank data of Resource Rents % in GDP and GDP per capita growth.

Figure 2: Smoothed Median Contract Intensive Money for 87 Developing and Emerging Economies, 1991-2011

Source: Own Calculation based on IMF World Economic Outlook Database

The Research Hypothesis and Data

The main hypothesis of this section is that the interaction between institutional quality and natural resource endowments impacts economic growth. We suggest that the natural resource abundance bare economic development only with qualitatively good economic institutions. Otherwise, if institutions demonstrate inefficiency in coping with conflicts, violence and rent-seeking activities, the natural resource abundance is negatively associated with economic growth.

The basic econometric model for the suggested interaction effect resource endowments with institutional quality on economic growth is defined as follows:

, (1)

where is the average yearly GDP per capita growth, is the vector of control variables such initial income per capita, foreign direct investment. The short description of the data used in research is presented in Table1. The dependent variable is , average growth rate of GDP per capita between 1991 and 2000 (first subsample), 2001 and 2010 (second subsample). is an average resource rents share in income for two subsamples - the measure of resource abundance and endowments and is the average measure of institutional quality for two subsamples again. To test the hypothesis that the resource endowments associated with good institutions drives economic growth, the interaction term - - is introduced. In line with the hypothesis suggested above, one would expect negative sign for coefficient implying resource curse phenomenon and positive sign for coefficient (commonly accepted finding) suggesting that good institutions determine economic growth, and positive sign for coefficient as well according to the hypothesis claimed in the broad literature (e.g. Brunnschweiler, 2007; Boschini et al., 2007). In case > │ │is documented, it implies that resource curse is been reversed. The key hypothesis of this study is to show that resource curse impact on economic growth appears through institutions. It goes in line with Sachs and Warner (1995) who have suspected it but ignored the hypothesis, preferring the Dutch Disease explanation. Thus, the main hypothesis of the study is that the rent-seeking activities deteriorate the institutional quality and undermine economic growth.

Contract-Intensive Money (CIM), proposed by Clague et al. (1999), is used as a proxy for institutional quality in this study. They suggest that society accumulate potential gains from business activities and potential trades boosted by contract enforcement and property rights. And the level of those potential gains that society can capture could be approximated by relative money in use. Contract-intensive money, suggested by Clague et al. (1999), is defined in this study as follows:

Inst = ,

where is a money supply including currency and deposits, C is a currency in circulation. If Inst (CIM) is a good proxy for contract enforcement and property rights in a broad sense, it should be a good indicator of the following characteristics of government’s role in the economy: (a) third-party enforcement for transactions and trades that cannot be realized otherwise; (b) it may function as an intermediary institution that links breaches of contract; (c) it may also establish the set of rules and arrangements in a way that private actors can themselves form formal groups (e.g. trade associations); (d) the government guarantees the peace. Obviously, one must be careful in choosing the approximate for institutional quality. The standard proxy variables employed in a broad literature of resource curse are indices such as ICRG, BERI and BI ratings (pioneered by Knack and Keefer, 1995; Mauro, 1995) and Worldwide Governance Indicators, suggested by Kaufmann et al. (Daniel Kaufmann, Aart Kraay and Massimo Mastruzzi (2010). "The Worldwide Governance Indicators : A Summary of Methodology, Data and Analytical Issues". World Bank Policy Research Working Paper No. 5430 ). However, there is a potential danger in using such indicators. The potential danger arises from the fact that these subjective measures could be misled by outcomes. For instance, the evaluators may report that the governance is good in times the countries demonstrate good and sound economic performance. Using CIM bares also potential risk and danger, if the measure is idiosyncratic and has little to do with contract enforcement and property rights. Clague et al. (1999) provide case studies of several countries and they find that CIM is a good measure for institutional quality, even though some countries examples show that there are cases of idiosyncratic measure.

As an illustration of the institutional quality variables relationship and other economic indicators the correlation matrix between Inst, indicators of Governance by Kaufmann et al. (2009) and other macroeconomic variables is provided in Table 1A. All indicators and measures are averages during 2001-2010. The correlation of Worldwide Governance Indicator (WGI), estimated as weighted average of its aggregate indicators, and its aggregate indicators of six broad dimension of governance such as Voice and Accountability (VA), Political Stability and Absence of Violence (PA), Government Effectiveness (GE), Regulatory Quality (RQ), Rule of Law (RL), Control of Corruption (CC) with CIM is illustrated to examine the adequacy of the latter as institutional quality variable is demonstrated. Among all other indicators CIM is mostly correlated with Government Effectiveness, Regulatory Quality and Rule of Law. It is highly correlated with other aggregates of WGI such as VA, PA and CC. Finally, CIM has a good correlation with WGI as well. We can see from the Table 1A that CIM is positively correlated with GDP per capita growth rate (Y). It is also detected that Resource Rents (RR) is not correlated with GDP per capita and other macroeconomic indicators. However, RR is positively correlated with all institutional quality indicators. On the other hand, RR is not correlated with other macroeconomic indicators such as foreign direct investment (FDI) and export share in GDP (Exp).

Table 1: Descriptive Statistics of the Main Variables Used in Regression Models

Dependent Variable	Definition	Source	During 1991-2000	During 2001-2010
Obs.	Mean	Std. Dev.	Obs.	Mean	Std. dev.
	Income per capita growth rate (annual %)	World Bank database		1.405	2.145	2.662	2.083
Inst	Contract-Intensive Money	World Bank database		0.538	0.212	0.528	0.222
FDI	Foreign Direct Investment, net inflows (% of GDP)	World Bank database		2.640	3.639	4.120	4.373
RR	Total natural resources rents are the sum of oil rents, natural gas rents, coal rents (hard and soft), mineral rents, and forest rents (% of GDP)	World Bank database		8.742	11.530	12.064	15.369
WGI	Weighted Average of Worldwide Governance Indicators	Kaufmann et al. (2010)	-	-	-	-0.370	0.713
GE	One of the six Worldwide Governance Indicators: Government Effectiveness	Kaufmann et al. (2010)	-	-	-	-0.377	0.717
RL	One of the six Worldwide Governance Indicators: Rule of Law	Kaufmann et al. (2010)	-	-	-	-0.477	0.702
lgdpea70	Log of Initial income per capita*	Sachs and Warner (1995)		7.878	0.756	7.878	0.756
EFrac	Ethnic fractionalization index*	Alesina et al. (2003)		0.539	0.246	0.539	0.246
Language	Linguistic fractionalization index*	Alesina et al. (2003)		0.479	0.317	0.479	0.317

*-The same data for separate subsamples

Table 2: Correlation between CIM, Kaufmann et al. Institutional Quality Indicators and Macroeconomic indicators, averages of 2001-2010

CIM

FDI

Exp

WGI

CIM

0.48***

1.00

0.28**

0.52***

1.00

0.67***

0.57***

0.63***

1.00

0.64***

0.62***

0.65***

0.92***

1.00

0.58***

0.54***

0.73***

0.93***

0.90***

1.00

0.51***

0.54***

0.70***

0.91***

0.88***

0.93***

1.00

0.23**

0.05

-0.01

0.21*

0.14

0.18

0.13

1.00

-0.25**

-0.37***

-0.13

-0.24*

-0.25**

-0.23**

-0.24**

0.11

1.00

FDI

-0.07

0.04

0.18

0.14

0.17

0.29***

0.23**

0.15

1.00

Exp

0.28**

0.13

0.42***

0.56***

0.51***

0.50***

0.55***

0.06

0.14

0.52***

1.00

WGI

0.40***

0.51***

0.53***

0.65***

0.69***

0.67***

0.71***

-0.02

-0.31***

0.12

0.34***

1.00

*** denotes significance at the 1% level; ** at the 5% level; * at the 10% level

All variables are average estimates of the indicators for the period 2001-2010; Number of observations are 81

CIM – Contract Intensive Money; VA – Voice and Accountability; PA - Political Stability and Absence of Violence; GE – Government Effectiveness; RQ – Regulatory Quality; RL – Rule of Law; CC – Control of Corruption; Y-GDP per cap growth; RR-Resource Rents; FDI – Foreign Direct Investment; Exp – Export share in GDP; WGI – Worldwide Governance Indicators Weighted Average Index of Institutional Quality