Defined by North (1994) as the “rules of the game”, over the last few decades, many scholars have sought to understand whether quality institutions can alleviate the Resource Curse (the idea that natural resource abundance hinders rather than promotes economic growth). However, with this focus on quality, few papers have addressed the question of institutional type, and its Curse mitigating properties. This paper, via utilising the associated data, contributes towards filling this gap. Using the Varieties of Capitalism framework, we test whether certain institutional typologies possess the ability to mitigate the Resource Curse and perhaps even turn it into a blessing. Specifically, Rougier and Combarnous’ cluster analysis, “The Diversity of Emerging Capitalisms in Developing Countries” (2017), whereby nations are assigned various institutional typologies is used to create our primary (dummy) independent variables of interest in this study. The remaining control variables essential for economic growth analysis are collected via the World Bank and Polity datasets.

The main independent variables of interest are those that represent which institutional typology a nation belongs to. Consequently, five dummy variables are created: Liberal, Coordinated, Informal, Paternalistic and Idiosyncratic (columns B-F in the dataset). These variables take the value of 1 when a nation belongs to the relevant cluster and 0 if not.

Regarding our measure of resource abundance, this paper creates a Resource Abundance variable (column J in the dataset) based on the World Bank’s measure of total resource rents as a percentage of GDP. Such a measure defines total natural resource rents as “the sum of oil rents, natural gas rents, coal rents (hard and soft), mineral rents, and forest rents.”

The Resource Abundance variable, as well as any other variable aside from the dummy variables that are not restricted to a single year is, for any given nation, an average taken from 1996-2017. The end date is chosen because, at the time of writing, this is the latest year provided by the World Bank for the aforementioned measure of natural resource rents. The start date is chosen because this is the earliest year the World Bank provides the data that is used to create our measure of institutional quality.

Four of the six World Bank Worldwide Governance Indicators are utilised as controls in this study. These are: Control of Corruption (column P in the dataset), Rule of Law (column Q in the dataset), Government Effectiveness (column N in the dataset) and Regulatory Quality (column O in the dataset). Additionally, an aggregate measure of Institutional Quality is calculated (column R in the dataset). This is done by, for a given nation, averaging the abovementioned four measures individually between 1996-2017, before finally averaging this calculated figure.

Three additional baseline controls are included. The data for each arises from the World Bank and are: Openness (an average from 1996-2017 of “the sum of exports and imports of goods and services measured as a share of gross domestic product”), Investment (an average from 1996-2017 of “the total value of the gross fixed capital formation and changes in inventories and acquisitions less disposals of valuables for a unit or sector” as a percentage of GDP) and Initial Income (the log of 1996 GDP per capita). Each variable (columns K, L and AH, respectively in the dataset) is derived from World Bank data.

Finally, our dependent variable of interest, Growth (column M in the dataset), is a measure of GDP per capita provided by the World Bank, averaged from 1996-2017. Specifically, this measure captures the “annual percentage growth rate of GDP per capita based on constant local currency… based on constant 2010 U.S. dollars… without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources”. For all variables derived from averages, missing data was resolved via a method of average imputation whereby said missing data was replaced by the period average.

Further variables are included in the dataset to test the robustness of results. These are:  measures of ethnic (column G in the dataset), linguistic (column H in the dataset) and religious (column I in the dataset) fractionalisation as defined by Alesina, et al (2003), a dummy variable (column S in the dataset), arising from the Polity IV (2019) dataset surrounding whether a nation experienced a regime transition or violent change during the sample period (Change), the log of nations’ population  in 1996-column T in the dataset- (Population), the sample period average of nations’ mortality rate-column U in the dataset (Mortality), government consumption as a percentage of GDP-column V in the dataset- (Government Consumption) and mineral rents as a percentage of GDP-column W in the dataset (Mineral Abundance). Additionally, legal origin dummies of Socialist, French, German and Scandinavian (columns Y-AB in the dataset) as defined by La Porta, et al (1999) where the omitted category is English (column X in the dataset) , as well as the regional dummies of  North America, Central and South America, Africa and Middle East and Asia and Oceania (columns AD-AG in the dataset) ,where the omitted category is Europe and Central Asia (column AC in the dataset), are included. Aside from the regional dummies, unless stated otherwise, all of the above variables are collected via World Bank datasets.

Where any of the metrics' data is unavaible for a nation, the relevant cell is presented as "N/A"

To replicate this dataset, interested parties will need access to the following sources of data:

1) The World Bank

2) Polity IV

3) Rougier, E. and Combarnous, F., 2017. The Diversity of Emerging Captialism in Developing Countries: Globalization, Institutional Convergence and Experimentation. Cham: Palgrave Macmillan (p. 111).

4) La Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R., 1999. The Quality of Government. Journal of Law, Economics and Organization, 15(1), pp. 222-279 (pp. 268-276 specifically).

Additionally, it is noteworthy that in calculating averages, missing data was resolved via a method of average imputation whereby said missing data was replaced by the period average.

To reanalyse the data, one can take the following steps:

1. Create interaction terms between institutional typology variables and the Resource Abundance variable as well as interaction terms between institutional quality variables and the Resource Abundance variable. The latter of these acts as a control variable.
   1. This is essential to test this paper’s hypothesis that it is institutional typology’s interaction with resource abundance that impacts economic growth.
2. Using the other control variables outlined in the above “Methods” section, create OLS multiple linear regression models of the following form:
   1. Y_i = β₀+ β₁Q_i +β₂W_i + β₃X_i +β₄Z_i +ε_I
   2. Here, Y_i is our dependent variable, Growth, Q_i is a vector of our included institutional type variables, W_i is a vector of our included institutional type variables’ interaction with Resource Abundance, X_i is Resource Abundance, and Z_i is a vector of our included controls.
   3. Researchers can vary the omitted institutional typology variable depending on differing institutional type comparison desires.
3. To test the robustness of results, add to the regression, both individually and collectively, the variables outlined above in the “Methods” section intended for such purpose, paying attention to whether the independent variables of interest change sign, magnitude or statistical significance.