The Peace Level of Nations: An Empirical Investigation into the Determining Factors

ABSTRACT

The escalated violent conflicts and political upheavals in many developing countries have emphasized the pertinence of examining the multifaceted nature of conflict, and the various strategies that bring about reasonable degrees of peace. This paper examines the effects of national economic and socio-economic factors on national peace level, and on the corresponding elementary indicators. The empirical analysis is implemented through a panel dataset, using different econometric methodologies. The basic results underline that countries characterized by higher economic development levels, open trade systems, more educated population, and democratic systems rest on higher national peace levels. Meanwhile, countries that experience higher levels of income inequality and that are endowed with natural resources tend to be less peaceful. Also, the positive impacts of international alliances/regional blocs on national peace are mainly expressed through their promoting economic effects rather than through their aggression-deterrence properties. The empirical analysis shows that the effects of national economic and socio-economic factors on the elementary indicators exhibit considerable variations in magnitude and significance. Hence, an exclusive examination of the effects of these variables on the overall peace index would conceal significant differences across the elementary indicators, which should be accounted for when analyzing national peace and developing peace-promoting strategies.

KEYWORDS:

• Peace
• conflict
• global peace index
• elementary indicators
• panel data

JEL Classification:

• D74
• F51
• H56
• Q34
• C23

Acknowledgments

The authors would like to thank the anonymous reviewers and the editor, Prof. Khusrav Gaibulloev, for their constructive and insightful comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Peace does not essentially imply the non-existence of any form of conflict, but it entails the absence of violence in all forms. Hence, it mainly stems from non-violent interaction and appropriate conflict management (Galtung 1996).

2. De Groot (2010) also finds that conflicts may have positive spillovers for non-contiguous neighbouring countries, which may benefit from some diversion of economic activities.

3. Inglehart and Welzel (2009) argue that, for the economic development to sustain higher peace levels, democratization in the presence of effective democratic institutions must prevail.

4. Cramer (2005) underlines that the theoretical basis of the negative effect of income inequality on national peace level is not always matched with empirical outcomes.

5. Fearon and Laitin (2003) indicate that the common measures of income inequality (e.g. Gini indicator) may not be reflective of political and ethnic grievances. There is a strand of the literature that analyzes inequality between different groups (e.g. ethnic groups) vis-à-vis inequality across individuals (Stewart 2000; Bartusevicius 2014).

6. Kilchevsky, Cason, and Wandschneider (2007) provide evidence that Middle Eastern countries enjoy more peaceful relationships with countries with which they are economically inter-dependent.

7. Grieco (1988) argues that trade might not act as a significant conflict deterrent when short-run gains outweigh long-run gains. Murshed and Gates (2005) indicate that the effect of trade on conflicts varies with the type of traded goods. They suggest that trade in military goods would generate higher tendencies to promote peace due to the strategic relevance compared to other types of traded goods (e.g. consumption goods).

8. Various models of economic growth (Solow 1957; Lucas 1988; Bils and Klenow 2000) stress the importance of education for human capital accumulation to uphold sustainable economic growth. The relationship between education and economic growth is critical for the maintenance of national peace since education per se in the absence of economic growth may not be enough to prevent violence (Hoffman 2011).

9. Agbor (2015) finds that primary education in Sub-Saharan Africa (SSA) has contributed in alleviating armed conflicts.

10. According to Gartzke, Li, and Boehmer (2001), democratic states often share more peaceful relationships with other democracies vis-à-vis non-democratic states.

11. Collier and Hoeffler (2004) note that lootable natural resources (i.e. those that are easily extracted) are easy targets for rebel takeover, and that they constitute significant sources for start-up funding of rebellion movements.

12. Auty (2001) explains that the exposure to price shocks is an important factor that lies behind the prevalence of economic instability in resource-dependent countries. Also, Le Billon (2001) argues that predictions about resource revenues do not necessarily encompass potential price shocks. Hence, fluctuations in resource prices would impede effective fiscal policies, and would increase income inequality, poverty, and grievance.

13. There is a difference between adjusting GDP for inflation and adjusting GDP for PPP. The variability in the GDP statistics expressed in US dollars at market rates is primarily derived from exchange rate variability. Hence, when using GDPC adjusted for inflation, the results may reflect the responses of GPI to actual changes in the value of final goods and services produced or to exchange rate variability. In this context, we note that the exchange rate uncertainty might be a predictor of peace. We thank an anonymous reviewer for this note.

14. Data on total natural resource endowments are limited. The correlation coefficient between natural resource exports and the available natural resource endowment data stands at a high value of 0.85.

15. The decomposition of the parameters into within-group estimator and between-group estimator can be illustrated as (Greene 2012): $b^{within}={\left[S_{xx}^{within}\right]}^{-1}{S}_{xy}^{within}$ where $S_{xx}^{within}$is the within-group moment matrix, and $S_{xy}^{within}$ is the matrix weighted average of the within-group estimator, and $b^{between}={\left[S_{xx}^{between}\right]}^{-1}{S}_{xy}^{between}$, where $S_{xx}^{between}$ is the between-group moment matrix, and $S_{xy}^{between}$ is the matrix weighted average of the between-group estimator.

16. We thank an anonymous reviewer for this suggestion.

17. The regressions can be alternatively carried out with the original variables.

18. Consider the following empirical model where observations are stacked by cluster group: $y_g=x_g^{\prime }\beta +{\epsilon }_g$ with $g=1,\dots ,G$. Then, the corresponding clustered variance is determined as: $V\hat{a}r\left(\hat{\beta }\right)={\left[X^{\prime }X\right]}^{-1}\left[{\sum }_{i=1}^G{x}_g^{\prime }{\hat{\epsilon }}_g{\hat{\epsilon }}_g^{\prime }{x}_g\right]{\left[X^{\prime }X\right]}^{-1}$ The use of clustered standard errors assumes that correlations do not prevail between clusters.

19. The estimated coefficients on NAFTA in columns (2) and (4) are not statistically significant. They become, however, positive and gain statistical significance when taking out Mexico from the NAFTA variable.

20. It should be noted that this estimate could reflect the net outcome from the different counteracting effects of education on the incarceration index.

21. Interestingly, the results show that democracies tend to increase the external conflicts fought (ECF) index. This outcome can be explained through the military interventions of developed democratic nations against some developing nations to overthrow dictators, remove governments with hostile political ideology, and/or exploit their resources. In this context, it is commonly observed that democracies are less likely to fight each other, but are more likely to get engaged in conflicts with non-democratic nations (Ray 1998).

22. These estimates could reflect the net outcomes from the different counteracting effects of resource endowments on elementary indicators.