DOES REALLY ECONOMIC FREEDOM MATTER FOR GROWTH IN SOUTH ASIA? EMPIRICAL EVIDENCES FROM PRE-ECONOMIC CRISES AND POST-ECONOMIC CRISES PERIOD

Muhammad Nadeem¹⁺ --- Yang Jun² --- Tahseen Akhtar³ --- Wanlu Dong⁴ --- Momna Niazi⁵

^1,2,5School of International Trade and Economics University of International Business and Economics, Beijing, PR China
³Professor, School of International Trade and Economic, University of International Business and Economics, Beijing, PR China
⁴Research Scholar, University of Sargodha, Punjab, Pakistan

Keywords:Panel data analysis, Economic freedom, Economic growth., Redundant fixed effects, Fixed effects, South Asia.

JEL Classification:B-15, C-33, O-10, 0-40.

ARTICLE HISTORY: Received:8 October 2018 Revised:13 November 2018 Accepted:16 December 2018 Published:2 January 2019.

Contribution/ Originality:This study contributes in existing literature by checking the role of economic freedom and its individual components for economic growth of South Asia without considering the role of global economic crises of 2008 and then adds up the post crises period into analysis. The results of Fixed effects test on two different models shows that global economic crises severely deteriorates the impact of economic freedom and one of its components (monetary freedom) tends with negative impact on economic growth on selected countries.

1. INTRODUCTION

South Asia is the fastest growing region in the world with average GDP growth 5-7 percent since during last two decades and it is projected with 7.1 percent by 2018. After global financial crises of 2008 several decisions were made by South Asian Association for Regional Cooperation (SAARC) countries in its summit of 2010. Secretary General Fathimath Saeed stated that under the theme of “Building Bridges” and more liberalization among member countries would be source of closing of space between intent and implementation of regional policies. President Mohamed Nasheed, chairperson of 17th SAARC summit also addressed areas of cooperation in which there is need of intense progress should take place, such as trade liberalization, more economic integration and integrity within and among member states.

Taking consideration of the above facts, it is necessary for the member countries to restore economic liberalization and governments must be careful about economic freedom if they seeks strengthen and revival of their economies. As focus of current study is to check empirical relationship between economic freedom and economic growth is related to the literature as observed earlier, but our contribution to the literature is that this study is checking the impact of economic freedom index (EFI) and individual components of economic freedom index on growth for both pre-crises period and overall observed period for South Asian countries.

Since early 1990’s, South Asian countries getting transformation from traditional economies to economically, politically and socially free economies (Khan, 2012). With this freedom, these countries get significant achievements with 5.5 percent overall economic growths during last two decades. The objective of South Asian countries is to attain consistency in economic growth. Different steps taken by these countries during 1990-2015 reflect that their concerns about economic liberalization are very serious. 

The concept of economic freedom is not new and has been discussed by different economists in different ways since economic theory is properly documented from the period of Adam Smith (Corbi, 2007; Ismail, 2010). Economic freedom does not have definition but a subjective value judgment. So economic freedom is explained in different ways in literature, but this study used “Economic Freedom Index” defined by Fraser Institute (Gwartney et al., 2015). Numerous studies conducted investigation to check possible linkages between economic freedom and economic growth, and most of the studies found that there is positive and significant impact of economic freedom on  growth (Barro, 1996; De Haan and Siermann, 1998; Cebula and Ekstrom, 2009; Khan, 2012; Bujancă and Ulman, 2015; Acikgoz et al., 2016). Rest of the roadmap of this study is based on literature review, data and methodology, results and discussion and conclusion.

2. LITERATURE REVIEW

The relationship between different individual measures of economic freedom and growth is not as clearer as different scholars suggested in different studies. It is concluded that there is strong correlation between various individual measures of economic freedom and growth across different selected nations (Spindler and Miyake, 1992; Vanssay and Spindler, 1994; Dawson, 1998). There are two major concerns related to economic freedom. First, if economic freedom is normal good, then an increase in income/wealth may increase its demand, in this case government is bound to provide more freedom on demand by individuals, it makes sense that increase in economic growth causes a higher degree of economic freedom when measured in a single index and vice versa is typical as discussed in most of the literature.

Second concern is related to its different measures and their implications across developed and developing nations. Typically most of the literature focused on first concern and second concern is more interested and is a comprehensive measure of different economic policies, especially for rapid growing countries (Bangladesh, India, Nepal, Pakistan and Sri Lanka). Empirically the coefficient of correlation found robust by different weighting methods of components of economic freedom and can create biasness if wrong weighting is used (Hanke and Walters, 1997).

Economic freedom can be used to explain cross-country differences in their economic performance. Although many of the studies used different economic freedom indexes but most of them concluded that there is significant positive impact of economic freedom on growth while taking a measure of overall index of economic freedom (De Haan and Siermann, 1998; Heckelman, 2000; Ali and Crain, 2001; Gwartney et al., 2006; Cebula and Clark, 2012). Contrary to the previous studies, there are some studies who found that there is no positive relationship between economic freedom and growth. Five out of 14 measures of economic freedom indexes measured by Fraser Institute effect growth adversely over the period of 1980-1992 (Heckelman and Stroup, 2005). Carlsson and Lundström (2002) concluded that some components of economic freedom are insignificant with growth; some of them are negatively related with growth in 75 selected countries for the period of 25 years.

3. DATA AND METHODOLOGY

Sample of our study comprises 5 South Asian countries for the period of 1990-2015 and data is taken from the Penn World Table v9.0, World Development Indicators (WDI) and datasets for the economic freedom is retrieved from the Fraser Institute¹ (Gwartney et al., 2015).

The core of economic freedom concept is “personal choice, property rights protection and freedom of exchange” (Gwartney, 1970). The economic freedom index is made up with five components namely size of government (SG), legal system and security of property rights (PR), Sound money (MF), freedom to trade internationally (TF) and regulation (REGU) which all together have further 24 sub-indicators. Each sub-indicator is rated between 0 and 10 and the rating of each component is based on the average of its sub-indicators and overall economic freedom index score is the average of the five components. Detailed description of the individual components of economic freedom is given in following table.

Table-1. Detailed Description of Component of Economic Freedom

Source: Fraser Institute

GPD per capita (constant 2010 US$) is used as a proxy of growth following (De Haan and Sturm, 2000; Adkins et al., 2002; Gwartney et al., 2006) for selected country in this study. Data on foreign direct investment (FDI) is taken in net inflows at current US$, Household final consumption expenditure is taken at constant 2010 US$ to control the demand side effects and “K” is gross fixed capital formation at constant 2010 US$, are taken from World Development Indicators (WDI). Data on human capital index (HC) is taken from Penn World Table v9.0 and this index is based on average years of schooling and return to education following the studies of Psacharopoulos (1994) and Barro and Lee (2013). 

This study uses panel data over the period of 1990-2015 for 5 South Asian countries and we tried our best to take all of the countries of the region to avoid the problem of country selection bias but the data for Afghanistan, Bhutan and Maldives was not available at given sources. Following Panahi et al. (2014) measures of individual indicators of economic freedom and their importance are considered for the aggregated rating of the selected countries.

3.1. Model Specification and Estimation Method

Following the existing literature on economic freedom-growth relationship in selecting variables² , our model specification is as follows:

The explanatory variables are lagged by one year to avoid the problem of simultaneity. In extension of the above model we now turn to the model 2 with five categories of the economic freedom index following the analysis of Carlsson and Lundström (2002) and estimated model is as follows:

Descriptive statistics is considered for all of the variables and given in table 2. This study uses linear panel data estimation to estimate model (1) and model (2) and during review of literature it is noted that Panel Least square (PLS), Fixed-effects model (FEM) and Random-effects model (REM) are used for such kind of linear panel data estimations.

Table-2. Descriptive Statistics

Source: Author’s own calculations

According to Baltagi (2005) that panel data technique has some advantages that are why this study uses longitudal data technique. First, panel data can control the problem of heterogeneity, while time series and cross-section cannot do this for different units. Second, panel data is more precise, having more variability, less chances of co-linearity between different variables with more degree of freedom and efficiency; third, panel data is more favorable with dynamic of adjustment and it is also suitable to study more complicated behavioral models.

Table-3. Correlation Matrix

Source: Author’s own calculations

In order to examine the source of correlation, the results of correlation matrix are shown in table 3 and it is noted that capital stock and economic freedom index have notably high correlation of 0.89 and 0.61 respectively with GDP per capita. Apart from that some of the components of economic freedom have also relatively high correlation with GDP per capita such trade freedom (TF) and regulations (REGU) have at round 0.51 and 0.55 respectively and these coefficients of correlation affirms that some of the explanatory variables may be correlated. Moreover it is interesting that monetary freedom has negative correlation with GDP per capita with human capital (HC) and household final consumption expenditure (HFC).

Table-4. Country Ranking and Summary Index (1990 and 2015)

Source: Fraser Institute

3.2. Methodology

As a first step of this study we checked cross-sectional dependency among series. For this purpose we follow the methodology of Pesaran (2004) which investigates the normal approximation, where it is assumed that mean variance of the indicator is approximated at T-1 and it is denoted by NLM. The mean and variance of Lagrange Multiplier (LM) test are provided to test the bias adjustments and proposed test affirmed the standard normal distribution for the fixed time series dimensions (T) when cross-section dimension (N) tend to infinity and bias NLM test is consistent even though Pesaran (2004) CD test is inconsistent. The results of bias-adjusted NLM test show that it controls the size successfully with maintaining satisfactory power in panel data with exogenous regressors and normal errors with mean of factor loading is close to zero.

Before proceeding to the final estimation technique it is necessary to control the order of integration of our series to get unbiased estimation results. As panel unit root testing is derived from time series unit root testing so this study needs to consider asymptotic trend of time series dimension T and cross-sectional dimension N (Nell and Zimmermann, 2011). For our study we used the approach of Augmented Dicky Fuller Fisher, Phillips-Perron Fisher, Levin et al. (2002) and Im et al. (2003) as our panel unit root tests and denoted them as ADF-Fisher, IPS, LLC and PP-Fisher respectively.

It is noted from literature that panel data may have group effects or time effects or both of them and these effects can be either fixed-effects or random-effects. As it is the property of panel data that if panel is balanced with all existing cross-sections then we can expect that fixed effects model may work perfectly (Asteriou and Hall, 2007). So this study use Redundant Fixed Effects test to check whether Panel Least square or Fixed effects model will be appropriate for the analysis.  A fixed-effect model works on assumption that there are differences among intercepts across group effects or time effects and explores the possible relationship between outcome and predictor variables within household, firms and countries. This study used fixed effects model on the bases of the results of Redundant fixed effects test and the results are presented in table 7.

In order to examine the real effects of global financial crises on the influence of economic freedom index (EFI) and its components on GDP per capita growth, panel data regression was estimated both for the period of pre-crises (1990-2008) and post crises (1990-2015).

4. RESULTS AND DISCUSSION

Table 5 show the results of CDLM test which control the size and power in panel data with exogenous variables and normality of errors when mean of cross-section factor loading is near to zero (Pesaran, 2006). In our results, t-statistics is not significant for both of the models, which means that our cross-section units are independent and we can use first generation panel unit root tests to check the order of integration for estimation of our models.

Table-5. Bias-Adjusted Test (CDLM Test)

Source: Author’s own calculations

Table-6. Panel Unit Root Results

Source: Author’s own calculations

In table 6, one may see that panel unit root tests results show that log of GDP per capita and all the control variables are stationary at their first difference and economic freedom index (EFI) and its individual components are stationary at their level form. Results show that null hypothesis of unit root for log of GDP per capita and all control variables can be rejected at 5% level of significance after taking their first difference.

4.1. Redundant Fixed Effects Test

Redundant fixed effects test is applied to make a choice between Panel Least Square Model and Fixed Effects Model. It tests the hypothesis given below:

Ho: = Panel Least Square Model is appropriate.
H1: = Fixed Effects Model is appropriate.

If the value of F-statistics is greater than probability value then we reject the null hypothesis that Least Square Model is appropriate, thus it is in favor of Fixed Effects Model.

Table-7. Redundant Fixed Effects Test

Note: Author’s own calculations

The results of Redundant Fixed Effects test show that F-statistics value is highly significant. So our null hypothesis is rejected at 5 percent level of significance as 99.077232 > 0.0000. It strongly rejects the null hypothesis that Least Square Model is appropriate, thus the results are in support of Fixed Effects Model. It indicates the presence of strong individual effects (country-specific effects).

The results of the model 1 and model 2 after model adjustment to avoid the problem of multicollinearity are shown in table 8 and table 9.

Table-8. Fixed Effects Model Results (Dependent Variable: dlogGDPPC)

***, **, *statistically significant at 1%, 5% and10% level respectively. t-statistics are in parenthesis
Source: Author’s own calculation

The results obtained from the fixed effects shows that coefficient of economic freedom (EF) is positively significant in both pre-crises period (1990-2008) and overall observed period (1990-2015). It indicates that in pre-crises period, one unit increase in economic freedom index would increase economic growth by 0.41 percentage point. But the results after inclusion of the global crises period in model show that it deteriorates the relationship between economic freedom and growth. Similarly the relation between control variables and economic growth is also as our expectations as the coefficient of human capital (HC) indicates that one unit increase in human capital would raise the economic growth by 0.35 percentage point. The coefficient of household final consumption expenditure (HFC) is negatively significant during both pre-crises period and post-crises period. It indicates that one percent increase in household final consumption would decrease the growth by 0.90 percentage point and 0.03 percentage point respectively.  The coefficient of foreign direct investment (FDI) shows that one percent increase in foreign direct investment would raise economic growth by 0.33 percentage point and 0.20 percentage point respectively and the coefficient of gross fixed capital formation (K) is also significant and indicates that one percent increase in capital would increase economic growth by 0.10 and 0.13 percentage point respectively. Except gross fixed capital formation, the coefficients of all these variables have been significantly deteriorated after the inclusion of the global financial crises period into the model. 

Table-9. Additional Fixed Effects Model Results (Dependent Variable: dlogGDPPC)

***, **, *statistically significant at 1%, 5% and10% level respectively. t-statistics are in parenthesis
Source: Author’s own calculation

In Table 8, summary of the results for the individual indicators is presented which shows that size of government (SG) is positively significant both in pre-crises and with higher coefficient in post crises period which indicates that one percent increase in size of government (SG) would raise economic growth by 0.49 and 0.60 percentage point respectively. Property rights protection under a country’s legal institutions is essential for economic growth (Kaur, 2006). The effect of the coefficient of property rights also shows the same trend as size of government (SG) in both periods that one unit increase in property rights index increases the economic growth by 0.41 percentage points and 0.64 percentage point after the inclusion of global crises period into regression. Relationship between property rights protection (PR) and GDP per capita growth is due to performance of property rights institutions at micro level which is increasing since late 1980’s in Bangladesh, India, Pakistan and Sri Lanka (Fernandez and Kraay, 2005). The coefficient of monetary freedom (MF) reflecting the access to the sound money is statistically significant and its sign is positive in pre-crises period estimation results which indicates that an increase in the monetary freedom by one unit reduces the growth by 0.93 percentage point in pre-crises period and its effect deteriorates with the inclusion of the global crises period  with negative coefficient which means that one percent increase in monetary freedom would decrease the growth by 0.20 percentage point. An increase in trade freedom by one unit leads toward economic growth by 0.53 percentage point in pre-crises period and it increase to 0.75 percentage point in overall observed period analysis. These finding are in line with the study of Ivanović and Stanišić (2017) which found a positive and robust relationship between trade freedom (TF) and economic growth in both pre-crises period and observed period for new European Union member states. And finally the coefficient of regulations (REGU) is positively significant which means that one unit increases in regulation (REGU) index raises the economic growth by 0.12 percentage point in pre-crises period and 0.80 percentage point in overall observed period. However, four of the components of economic freedom (SG, PR, TF, REGU) are positively significant, coefficient of monetary freedom (MF) is negatively significant in post-crises period and with inclusion of crises period into the model 2, increases the value of the coefficients of other four components.

Apart from the significance of the coefficients of explanatory variables, value of R2 also deteriorates i.e. it is high (0.6488) in pre-crises period (1990-2008) and decreases to (0.5498) in post-crises period (1990-2015) in model 1 and model 2 also show the same results such as the value of R2 in pre-crises period (19990-2008) was high (0.6922) and decreases to (0.5648) in post-crises period. It shows that global financial crises of 2008 severely blurred the link between economic freedom and growth in South Asia. Most of the studies in previous literature about economic freedom-growth nexus show positive and robust relationship which is not endorsed by the empirical results of our study and to check the pre-crises effects and observed period effects of economic freedom index and its components with all these control variables on economic growth is purely our contribution to the literature.

Table-10. Result Summary of the Effects of Economic Freedom Components on Economic Growth

Source: Author’s own calculations

5. CONCLUSION

This study examines the relationship between economic freedom and growth in 5 South Asian countries. The real contribution of this study stems from the fact that very number of previous studies explored this relationship for South Asia. The primary focus of previous studies was to check the impact of economic freedom index and growth performance in South Asia. Adding to the literature, this study test the relationship between economic freedom index (EFI), its components and economic growth for the period of 1990-2015 covering the pre-crises and post-crises to evaluate the impact of economic crises  of 2008 on freedom-growth relationship.

The first hypothesis of this study was that economic freedom has positive relationship with economic growth and empirical findings supports this hypothesis over the period of 1990-2015. These results are compatible with variety of previous studies. The second hypothesis of our study was that individual components of economic freedom are positively correlated with growth and empirical findings does not support for all of the components such as monetary freedom (MF) is negatively correlated with GDP per capita growth in post-crises period. These findings are not compatible with prior studies for South Asian countries. The third hypothesis of our study was that global economic crises of 2008 blurred the relationship between economic freedom, its components and economic growth. The empirical findings of model 1 suggest that relationship between explanatory variables and GDP per capita growth is considerably deteriorated after inclusion the period of economic crises in to the model and model 2 suggest that one of the components of economic freedom index such monetary freedom changed its contribution toward growth in post-crises period. The findings suggest that South Asian countries were not maintaining the level of economic freedom and its financial system to face the global financial recession. This implies that relative focus of further research is moved toward the structural parameters of growth equation to absorb sudden economic shocks in these economies. Further research on economic freedom-growth nexus could be on the identification of those factors which influence the relationship between components of economic freedom and growth in South Asia.

REFERENCES

Acikgoz, B., A. Amoah and M. Yılmazer, 2016. Economic freedom and growth: A panel cointegration approach. Panoeconomicus, 63(5): 541-562. Available at: https://doi.org/10.2298/pan140520018a.

Adkins, L.C., R.L. Moomaw and A. Savvides, 2002. Institutions, freedom, and technical efficiency. Southern Economic Journal, 69(1): 92-92. Available at: https://doi.org/10.2307/1061558.

Ali, A. and W.M. Crain, 2001. Political regimes, economic freedom, institutions and growth. Public Choice, 19(1): 3–22.

Asteriou, D. and G.S. Hall, 2007. Applied econometrics. Palgrave Macmillan, Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York.

Baltagi, B., 2005. Econometric analysis of panel data. 3rd Edn., Chichester: John Wiley Sons, Ltd.

Barro, R.J., 1996. Determinants of economic growth: A cross-country empirical study. National Bureau of Economic Research (No. w5698).

Barro, R.J. and J.W. Lee, 2013. A new data set of educational attainment in the world, 1950–2010. Journal of Development Economics, 104: 184-198.

Bujancă, G.V. and S.R. Ulman, 2015. The impact of the economic freedom on national competitiveness in the main economic power centers in the world. Procedia Economics and Finance, 20: 94-103. Available at: https://doi.org/10.1016/s2212-5671(15)00052-0.

Carlsson, F. and S. Lundström, 2002. Economic freedom and growth: Decomposing the effects. Public Choice, 112(3-4): 335-344.

Cebula, R. and M. Ekstrom, 2009. Joint impact of the dimensions of governance and economic freedom on economic growth in OECD nations: An analysis with controls for budget deficits, and G8 status. Research in Applied Economics, 1(1): 1-10. Available at: https://doi.org/10.5296/rae.v1i1.186.

Cebula, R.J. and J. Clark, 2012. Lessons from the experience of OECD nations on macroeconomic growth and economic freedom, 2004-2008. International Review of Economics, 59(3): 231-243. Available at: https://doi.org/10.1007/s12232-012-0162-0.

Corbi, R.B., 2007. The components of economic freedom, income and growth: An empirical analysis. Economic Studies (São Paulo), 37(3): 515-545.

Dawson, J.W., 1998. Institutions, investment, and growth: New cross-country and panel data evidence. Economic Inquiry, 36(4): 603–619. Available at: https://doi.org/10.1111/j.1465-7295.1998.tb01739.x.

De Haan, J. and C.L. Siermann, 1998. Further evidence on the relationship between economic freedom and economic growth. Public Choice, 95(3): 363-380.

De Haan, J. and J. Sturm, 2000. On the relationship between economic freedom and economic growth. European Journal of Political Economics, 16(2): 215–241.

Fernandez, A.M. and A. Kraay, 2005. Property rights institutions, contracting institutions, and growth in South Asia: Macro and Micro evidence. The World Bank. Background Paper Prepared for the SAARC Business Leaders’ Conclave: South Asia Regional Integration and Growth, New Delhi, November. pp: 17-18.

Gwartney, J., 1970. Discrimination and income differentials. The American Economic Review, 60(3): 396-408.

Gwartney, J., R. Lawson, J. Hall, J. Gwartney, R. Lawson and J. Hall, 2015. Economic Freedom Dataset, Published in Economic Freedom of the World: 2015 Annual Report.

Gwartney, J.D., R.G. Holcombe and R.A. Lawson, 2004. Economic freedom, institutional quality, and cross-country differences in income and growth. Cato J., 24: 205.

Gwartney, J.D., R.G. Holcombe and R.A. Lawson, 2006. Institutions and the impact of investment on growth. Kyklos, 59(2): 255-273.

Hanke, S.H. and S.J. Walters, 1997. Economic freedom, properity, and equality: A survey. Cato Journal, 17(2): 117-156.

Heckelman, J.C., 2000. Economic freedom and economic growth: A short-run causal investigation. Journal of Applied Economics, 3(1): 71-91.

Heckelman, J.C. and M.D. Stroup, 2005. A comparison of aggregation methods for measures of economic freedom. European Journal of Political Economy, 21(4): 953-966. Available at: https://doi.org/10.1016/j.ejpoleco.2005.06.001.

Im, K.S., M.H. Pesaran and Y. Shin, 2003. Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115(1): 53-74. Available at: https://doi.org/10.1016/s0304-4076(03)00092-7.

Ismail, N., 2010. Income inequality, economic freedom and economic growth. European Journal of Economics, Finance and Administrative Sciences, 25: 144-157.

Ivanović, V. and N. Stanišić, 2017. Monetary freedom and economic growth in new European Union member states. Economic Research-Economic Research, 30(1): 453-463. Available at: https://doi.org/10.1080/1331677x.2017.1305803.

Kaur, S., 2006. Inequality, growth and economic freedom: Re-examining the role of government. RGICS Working Paper Series, No. 54: 1-30.

Khan, M.Z.S., 2012. Democracy and economic freedom: A static panel data analysis of South Asia. CEU Political Science Journal, 7(1): 32-50.

Levin, A., C.-F. Lin and C.-S.J. Chu, 2002. Unit root tests in panel data: Asymptotic and finite-sample properties. Journal of Econometrics, 108(1): 1-24. Available at: https://doi.org/10.1016/s0304-4076(01)00098-7.

Nell, C. and S. Zimmermann, 2011. Panel unit root tests term. Paper at the Department of Economics at University of Vienna, Phd Course: Panel Data. Lecturer: Prof. Dr. Robert Kunst.

Panahi, H., A. Assadzadeh and R. Refaei, 2014. Economic freedom and economic growth in Mena Countries. Asian Economic and Financial Review, 4(1): 105-116.

Pesaran, M.H., 2004. General diagnostic tests for cross section dependence in panels, CESifo Working Paper, No. 1229. Available from http://hdl.handle.net/10419/18868 .

Pesaran, M.H., 2006. Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica, 74(4): 967-1012. Available at: https://doi.org/10.1111/j.1468-0262.2006.00692.x.

Psacharopoulos, G., 1994. Returns to investment in education: A global update. World Development, 22(9): 1325-1343. Available at: https://doi.org/10.1016/0305-750x(94)90007-8.

Spindler, Z.A. and J.F. Miyake, 1992. Prospecting for the ‘homework’ measures of economic freedom: A summary. In S.T. Easton and M.A. Walker (Eds.), Rating Global Economic Freedom. Vancouver: The Fraser Institute. pp: 230-254.

Vanssay, D.X. and Z.A. Spindler, 1994. Freedom and growth: Do constitutions matter? Public Choice, 78(3-4): 359-372. Available at: https://doi.org/10.1007/bf01047764.

APPENDEX 1

Footnotes: