Defence Procurement

Image ID F5FJ70 © Sergiy Tryapitsyn/Alamy Stock Photo

AK-47 assault rifle.

Guns, Funds and Power: Defence Spending, Regime Type and the Implications for Canada

by Joshua Horlings

Print PDF

For more information on accessing this file, please visit our help page.

Second Lieutenant Joshua Horlings graduated from the Royal Military College of Canada in Kingston, Ontario, in May 2016 with a BA (Honours) in Military and Strategic Studies. He is currently completing his armoured phase training at CFB Gagetown in New Brunswick.

Introduction

An AK-47 assault rifle can be purchased for between $10-$300, depending upon where you are in the world. As a result of its relative cheapness and extreme durability, there are an estimated 75-100 million AK-47s worldwide—enough for one-in-every-60 of the world’s population.1 Assuming the entire world military expenditure for 2014 was spent on $300 AK-47s, it would be enough to purchase around 6 billion rifles—one for nearly every person on the globe. With global military expenditures rising yearly, the governments of the world are not only purchasing AK-47s but also tanks, fixed wing aircraft, attack helicopters, submarines, aircraft carriers and a plethora of other weapons platforms. Nearly every country has increased its military spending since 1990, including democracies—and that includes Canada. Democracies consistently spend huge amounts on defence while actually devoting only a small percentage of their GDP to defence. Autocracies on the other hand, devote larger percentages of their GDP to defence, while generally spending small amounts overall. This difference in spending habits between regime types prompts the question: does regime type determine the nature of a state’s defence spending?

Since autocracies require a large military to maintain control and to deter uprisings, it has already been established that they often spend more as a percentage of GDP on defence.2 However, if autocracies spend such a large percentage of their GDP on defence, why is it that they still spend so little overall? This leads to the first part of my hypothesis that autocracies will have a lower GDP growth rate and will spend a higher percentage of their GDP on defence than democracies. The rationale behind this hypothesis is that autocracies will have more of a focus on the military as a control measure, and less of a focus upon economic growth through trade or international cooperation.

Democracies, on the other hand, gain more from spending on social services and building trade, rather than spending on defence. This is because democratic regimes maintain power by keeping citizens happy through economic growth, and cannot resort to violent repression.3 Therefore, the second part of my hypothesis is that democracies will spend less as a percentage of their GDP, but more overall on defence. The rationale behind this is that since democracies focus upon economic growth, their economies are capable of huge output. Given this, my hypothesis is that autocratic regimes will spend a higher percentage of their GDP on defence than democratic regimes, but democratic regimes will have higher GDP growth than autocratic regimes, and higher overall spending.

DND photo BN2014-0070-08

A CF-18 Hornet escorts a CC-177 Globemaster III carrying Canadian Armed Forces members returning from the last mission in Afghanistan to the Ottawa International Airport, 18 March 2014.

Background

The initial question I had before writing this article was, why do democracies spend so much of defence, if democracies are supposed to be peaceful? However, by researching more into the democratic peace theory, I found that my understanding of that theory was wrong. Democracies are, in fact, often as violent as autocracies, and have fought many wars—just never with other established democracies.4 Kant built his argument for democratic peace on three elements, also known as the “Kantian Triangle of Peace.” These three elements build the basis for why democracies do not fight other democracies. They are “republican constitutions, ‘cosmopolitan law’ embodied in free trade and economic interdependence, and international law and organizations.”5 Essentially, democracies do not fight other democracies because there is no longer a distinction between “us” and “them.” Instead, the two states see a far larger benefit to economic relations and international co-operation, and a disadvantage in going to war. Furthermore, Kant argued that as democracy spread, defence spending would decrease since states would no longer need to fear their neighbors. Assuming that the way in which a government allocates GDP indicates importance, Kant theorized that as democracy spread, states would be less inclined to spend on defence. This is because he believed defence spending would be viewed as less important than social services, health care, and education. Based upon this reasoning, a state that spends a large percentage of GDP on defence would place a large importance on the military. Instead of military strength, a democratic government will pursue economic growth, and thus satisfy the second element of the Kantian triangle of peace through the emphasis on free trade and economic interdependence by democracies. As a byproduct of economic growth, however, democracies can get away with spending only a small percentage of their massive GDP on defence, since a small percentage can translate into billions of dollars. What these factors come down to is essentially a method of staying in power. An autocracy will use military strength to stay in power, whereas a democracy will use economic growth as a way of keeping voters happy and staying in power.

Aside from the democratic peace theory, I could also have examined defence spending in relation to external threat, and in relation to alliances or multilateral security. However the relationship between regime type and defence spending is less prevalent in these theories, and has already been examined by University of Mississippi professor, Jeff Carter and Pennsylvania State University professor, Glenn Palmer. Another option I considered was the use of case studies related to a few specific states. However, I believe using a large dataset and looking for statistical relationships will provide a clearer answer to my hypothesis. Seeing as the democratic peace theory is being considered “…perhaps the most influential domestic level framework in analyzing world politics, and the theories of how societies divide the world into ‘us’ and ‘them,’”6 I concluded that it would be the best theory for analyzing if regime type determines the nature of a state’s defence spending.

Image ID ERHGCE © Lebrecht Music and Arts Photo Library/Alamy Stock Photo

Immanuel Kant, the German Prussian philosopher, circa 1791 (G. Döbler).

The democratic peace theory is perhaps one of the most controversial and heavily researched theories in politics. Dating back to 1795, it was initially proposed by the German philosopher Immanuel Kant, who theorized that democracies are less likely to go to war than any other regime type.7 In major revisions by American international relations scholars Michael Doyle and Spencer Weart, the theory has changed and now proposes that democracies will fight a different regime type quite willingly, but will almost never wage war on another democracy.8 Many other studies have been conducted on the credibility of the theory, including those by prominent political scientists, such as Stuart Bremer and Robert Ivie.9 Rather than focus upon the democratic peace theory as a whole, I will focus more upon the literature pertaining to regime type and defence spending.

Research regarding the relationships between regime type and military spending has only really become a topic of interest over the past two or three decades. Perhaps this is due to the new understanding of the democratic peace theory, or as an effort to explain post-Cold War military spending trends. Regardless, the literature with respect to the determinants of defence spending and regime type is substantial. Studies into regime type and defence spending have found, for the most part, that Kant’s theory holds water. Democracies do, in fact, devote fewer resources to their militaries than autocracies. However, it is unclear if regime type is the strongest factor in determining this relationship.10 William Nordhaus, John Oneal, and Bruce Russett, a trio of American political scientists who have published a number of articles together on the Democratic Peace Theory, explore this question of causation in a study. Their study found that external security threats have a far larger implication on defence spending than does regime type.11 Other American political scientists have also examined the topic, including Carter and Palmer, who studied the role of external security threats in order to find patterns of defence spending among different regime types based on times of war and peace. This study focused more upon the economics of mobilization, regime type, and interstate war; finding that non-democratic regimes are more equipped to mobilize for war than democratic regimes. This perhaps indicates that autocratic regimes place a higher importance on the ability to use their military, and thus are willing to invest more.12 In terms of a comparison between regime types and percentage of GDP spent on defence, a study by a British academic, Jennifer Brauner, found that democracies do, in fact, spend less than autocracies as a percentage of GDP and that there was a causal relationship between regime type and military spending. The data examined in this study, however, did not go beyond the year 2000, and unlike my article, did not look to find a relationship with GDP growth. Brauner also mentions in her literature review that her study is the first that she is aware of to examine the causational relationship between regime type and defence spending as a percentage of GDP.13

In terms of literature regarding regime type and GDP growth, there is also a substantial body of work. Unlike literature regarding regime type and defence spending however, the studies in this area are far less conclusive. A 1993 study by the respected and award winning political scientist, Adam Przeworski, determined that “…we do not know whether democracy fosters or hinders economic growth. All we can offer at this moment are some educated guesses.”14 Similarly a 224-page study by Taiwanese political scientist Dr. Chin-en Wu on the topic found that regime type did encourage certain economic behaviors and that overall, democracies tended to be more prosperous. However, results were far from conclusive.15 In a comprehensive 529-page study on regime type and economic growth, the highly published Norwegian academic, Carl Knutsen, essentially concluded there are far too many factors at play and that there are too many outliers to draw any conclusions regarding regime type and economic growth.16 On the other hand, a study by prominent Bonn University professor, Erich Weede, found democratic regimes may have a detrimental effect on economic growth.17 There is currently a study underway at Stanford University investigating the relationship. However, the website description admits the difficulty in finding substantial results as, “…for every high-growth authoritarian regime like China there are authoritarian economic disasters like Zimbabwe.”18

Other studies of note include a study by University of Macedonia professor, Nikolaos Dritsakis, which looked to determine if there was a relationship between defence spending and economic growth, using the examples of Greece and Turkey. The study hypothesized that high defence spending would indicate good economic growth. However, it found no relationship between the two variables.19 This study varied from my own in that it used a more economically-focused view, and only examined two states, with no consideration of regime type. Another study referenced herein is a study by the prominent American political scientist, Ethan Kapstein, which examined the relationship between economics and security studies and built a case for a stronger emphasis upon economics within the field of security studies.20

Through this literature review, it is clear that my study fits within a very contemporary and changing field of research, with most prominent sources being written over the last ten-to-twenty years. It is clear, through examination of the existing literature, that regime type and defence spending have an established causal relationship. However, the relationship between regime type and GDP growth lacks conclusions. In terms of how I am connecting the three variables, I was unable to find any literature, thus giving me the opportunity to possibly extend this field of study.

DND photo HS2016-A060-003/Master Corporal Sebastian Allain

HMCS Fredericton performs a Passex with the Bulgarian frigate BGS Reshitelni (F13) and the Romanian frigate ROS Regina Maria during a patrol in the Black Sea as part of Operation Reassurance, 5 April 2016.

Discussion

In my analysis, I used data from 131 countries for the year 2013. The data I used covered five variables: regime type, percentage of GDP spent on defence, total government spending on defence, percentage of total government expenditure spent on defence, and GDP growth. I tested all the data in pairs in order to determine if there was any correlation. I then used the Pearson Product Moment Correlation Coefficient Table to analyze the results. By doing this, I was able to determine whether any correlation relationship I found was significant, and if so, to what degree it was significant. I chose the method of statistical analysis since it gives not only an indication if a relationship exists, but also how strong the relationship is. As well, since my topic is rooted in statistics, it made the most sense out of all the quantitative methods. The inherent difficulty in any relationship is differentiating correlation and causation. In this article, I will only be looking for correlation, and will be discussing any given relationship without searching for causation.

The data I used for regime type was taken from Freedom House’s Freedom in the World country ratings.21 The data is measured in two categories on a scale from one to seven. The two categories are Political Rights and Civil Liberties, and together, the two categories are combined to rank states as Free, Partially Free, and Not Free. For my research, I used only the raw data for political rights. I chose to only use the one set of data, since the two categories reflected essentially the same scores, and correlating with both Political Rights and Civil Liberties would have been redundant. I used the Freedom House’s data since the seven point scale provided a simple data set, yet still included enough variation for distinct score differences to exist between regimes. The scale ranked the most free states with a score of one and the least free states with a score of seven. Through these scores, I determined which states could be considered democratic or autocratic.

The data for percentage GDP spent on defence, total defence spending and percentage of government expenditure dedicated to defence was all taken from the Stockholm International Peace Research Institute’s (SIPRI) data set regarding armaments, disarmaments, and security.22 As mentioned earlier, the variable of percentage GDP spent on defence is important, because it indicates the degree of necessity placed by a regime upon the military. The variable of total defence spending indicates a state’s military capacity, and a state’s economic capability for output. I also included the variable for percentage of a government’s budget dedicated to defence as a second piece of data, indicating the necessity placed upon the military by a regime. The difficulty in researching the topic of defence spending is that data is sometimes unavailable or unreliable for many states. As a result, a number of states had to be removed from the data that I used, since no defence spending data was given, most of these states being autocratic regimes. As well, some of the data included was either created through SIPRI estimates, or highly uncertain. These data scores are indicated in blue and red respectively on the data table. In order to analyze the data further, I also conducted all the correlations in two different sets; one with the USA included and the other without. I chose to do this since the USA represents a huge outlier in overall defence spending, as it spends as much on defence as the next nine states in succession.

The fifth variable is GDP growth, and it was retrieved from a dataset provided by the World Bank.23 The importance of GDP growth indicates whether or not a state is growing economically, and by how much it is growing. The reason I chose this variable was based on my hypothesis that democracies will have larger economic growth than autocracies and that democracies will use this large growth as a method to stay in power. The reason I chose GDP growth rather than overall GDP was because I believe GDP growth gives a better view of a state’s economic health. GDP growth indicates that citizens’ standard of living is increasing and that citizens are employed, meaning, in accordance with my hypothesis, that they were satisfied with the government in power. For example, a state such as Greece that has experienced negative GDP growth over the past few years, has also experienced huge political turmoil. The dataset from the World Bank was fairly comprehensive. However, it lacked data for a few states that had to be removed from the dataset as well.

  Regime Type % GDP on Defence Total Defence Spending % Govt Spending on Defence GDP Growth
Regime Type 1.00        
% GDP on Defence 0.43*** 1.00      
Total Defence Spending -0.07 0.17* 1.00    
% Govt Spending on Defence 0.49*** 0.89*** 0.13 1.00  
GDP Growth 0.38*** 0.03 -0.07 0.17* 1.00
***p<0.01 **p<0.05 *p<0.10

Author

Figure 1 – Statistical Analysis Results for Regime Type, Percentage of GDP Spent on Defence, Total Defence Spending, Percentage of Total Government Spending and GDP growth. (Dataset includes USA)


Figures 1 and 2 show the correlation data found through a series of data correlations. The data used in Figure 1 included the USA where as Figure 2 did not do so. Between Figure 1 and 2, there are a number of differences. However, they are less dramatic than I expected. Due to the overall similarity between the two sets of results, I used the data from Figure 1 in constructing any graphs, since it includes the complete data set. I will, however, discuss Figure 2 later in the article.

  Regime Type % GDP on Defence Total Defence Spending % Govt Spending on Defence GDP Growth
Regime Type 1.00        
% GDP on Defence 0.45*** 1.00      
Total Defence Spending 0.06 0.20** 1.00    
% Govt Spending on Defence 0.50*** 0.89*** 0.17* 1.00  
GDP Growth 0.38*** 0.04 -0.06 0.18** 1.00
***p<0.01 **p<0.05 *p<0.10

Author

Figure 2 – Statistical Analysis for Regime Type, Percentage of GDP Spent on Defence, Total Defence Spending, Percentage of Total Government Spending and GDP Growth (Dataset without USA)


The results reflected in Figure 3 show a strong relationship that supports the conclusions made by the Brauner study that autocracies do, in fact, spend a higher percentage of their GDP on defence. This finding, therefore, was not a surprise, and it provided support for the first part of my hypothesis that autocracies will spend a higher percentage of their GDP on defence than do democracies. There are a number of interesting outliers in these results, including Israel and the USA, that rank highest among the Level 1 regimes. Oman, Saudi Arabia, and Afghanistan were also outliers with ranks of first, second, and third respectively among the Level 6 regimes. Although not unexpected as outliers, the degree by which these states outrank their counterparts is interesting. Interestingly, there are also no outstanding Level 7 regimes.

Author

Figure 3 – Regime Type and Percentage of GDP Spent on Defence

Click to enlarge image

As expected in Figure 4, the variable for percentage of total government expenditure spent on defence and regime type reflected the same findings as percentage GDP spent on defence and regime type. However in this relationship, there were far more outliers and a slightly stronger correlation. As in Figure 3, states such as Israel, Oman, Afghanistan, and Saudi Arabia can be easily identified as the highest spenders in their respective categories. However, this relationship brought out other states not significant in Figure 3. This included prominent Level 4 regimes, such as Singapore at 20.5%, and Pakistan at 16.2%. Also prominent in the Level 5 category were Armenia at 15.8%, and Sri Lanka at 14.2%. Another interesting note, as in Figure 3, is the lack of high spending Level 7 regimes…

Author

Figure 4 – Percentage of Government Expenditure Spent on Defence and Regime Type

Click to enlarge image

As expected in Figure 5, the USA was an extreme outlier in overall defence spending, dedicating an astounding $600 billion in the examined year. Other prominent states included China, Russia and Saudi Arabia. Despite the fact that the USA spends more than the next top nine defence spenders combined, no significant correlation was found between regime type and overall defence spending. I found this surprising, since I expected democratic regimes to correlate with higher overall spending. This finding goes against my hypothesis that democratic regimes will be correlated with more spending. Even with the USA removed, no substantial correlation existed connecting regime type and overall spending in any direction.

Author

Figure 5 – Regime Type and Overall Spending

Click to enlarge image

As seen in Figure 6, the relationship between regime type and GDP growth did not support my hypothesis. Rather than democracies having the highest GDP growth, many had the lowest. Overall, the data was very widespread, and I was surprised, despite the wide spread, by the strength of the correlation found between autocratic regimes and GDP growth. This finding is the opposite of what I expected. The states with the largest GDP growth were Paraguay (Level 2), Liberia (Level 3), Kyrgyzstan (Level 6), and Ethiopia (Level 6), none of which were significant in any other category.

Author

Figure 6 – Regime Type and GDP Growth

Click to enlarge image

In the clearest results against my hypothesis, Figure 7 shows the lack of any substantial correlation between percentage of GDP spent on defence and GDP growth. If these results had supported my hypothesis, I would have expected to see a clear grouping slanting down from the top left corner to the bottom right. However, these results show no relationship.

Author

Figure 7 – Percentage of GDP Spent on Defence and GDP Growth

Click to enlarge image

Although my research did not support my hypothesis, it did offer some interesting insight into what determines the nature of a state’s defence spending. My findings that autocracies spend a higher percentage of their GDP on defence match the findings of Brauner, as well as those of Fordham and Walker, and Yildirim and Sezgin. My research also supported Nordhaus, Oneal, and Russett in the finding that external security threats play a very important role in determining defence spending. Although I did not statistically prove this support, by looking at the highest spenders, an interesting common denominator can be found in an external security threat. For example, states such as Israel, Saudi Arabia, Singapore, South Korea, and Armenia, which all have major external security threats, all placed high in both percentage of GDP spent on defence, and percentage of total government expenditure spent on defence. Similarly, states in the midst of warfare, such as Afghanistan, the USA, Pakistan, and Colombia, all spent a fairly large percentage of their total government expenditure on defence.

My research, however, did not support Fordham and Walker, and Yildirim and Sezgin in the findings that democracies devote fewer resources to defence than other regime types. That said, this discrepancy is due to the fact that both studies used percentage of GDP spent on defence as their measure for resource allocation. I find this problematic, since a country such as Oman can allocate a larger percentage than the USA, but allocate fewer resources, due to a huge difference in GDP size. Although autocracies devote a larger percentage of GDP and government expenditure, there was no indication they contributed more overall. The correlation for this data, in fact, proved very low, and was very heavily influenced by ‘heavy spenders,’ such as the USA and China. Even when I removed the USA and China from the dataset, only a small correlation of -0.10 existed in favor of more democratic regimes spending more overall. This finding surprised me, seeing as the top ten defence spenders account for a huge percentage of total defence spending, and they are all ranked Level 1 or 2 regimes, except China, Russia, and Saudi Arabia.

Another interesting finding was the absence of any prominent Level 7 regimes other than China. As the most autocratic, I assumed these states would have ranked high in percentage of GDP spent on defence and percentage of government expenditure spent on defence. However, they failed to stand out in any way. One possible reason for this is the lack of available data for these states. Seeing as the large majority of states removed from the dataset, due to lack of data, were Level 6 and 7 regimes, my results were undoubtedly affected. Although there were also Level 1 and 2 states with unavailable data, I can only assume that states such as Syria, Sudan, or North Korea will outspend states, such as Mongolia and Iceland.

Between the two data sets on percentage of GDP percentage spent on defence and percentage of government expenditure spent on defence, there was a very strong correlation of 0.89. Although similar, I found the data with respect to percentage of government expenditure spent on defence gave a far more varied and insightful display of the data. Percentage of GDP spent on defence is widely accepted as a primary measure for defence spending. However, based upon my results, I would argue that percentage of total government expenditure provides a better analysis and perspective on outliers and on states for which the military is of high importance.

In relation to the minimal impact of the USA on the data, I can only attribute this to the fact that the USA is only outstanding in its overall defence spending since it spends a low 3.8% of its GDP on defence, and an unremarkable 10% of its total government expenditure on defence.

In my hypothesis, I attempted to relate the nature of defence spending with a method for a regime to maintain power. Based upon my findings, I believe there is a relationship between autocracies maintaining power by spending on the military. However it appears as though there is no relationship between democracies, high GDP growth and low percentage of GDP spending on defence. Although many democracies spent only a small percentage of their GDP on defence, these states did not necessarily have high GDP growth. I believe this points to a fundamental error in my choice of variables. Although as mentioned in my methods section, I chose GDP growth as a variable since it better portrays economic health, it proved ineffective in gauging the health of a state as a whole. By using GDP growth, established democratic states such as the USA, France, Germany, and Canada stood no chance in ranking high since with huge existing GDPs, any GDP growth as a percentage is extremely hard to attain. On the other hand, tiny underdeveloped states, such as Liberia, Moldova, and the Democratic People’s Republic of Congo could experience huge rises in GDP growth while actually growing very minimally, due to tiny existing GDPs. As a result, some of the most underdeveloped and authoritarian states exemplified large GDP growth since GDP growth is based upon GDP size, and not actual output. Therefore, it would be interesting to change the variable of GDP growth to total GDP size and make the argument that it is overall GDP size that makes citizens happy and content with government, and not GDP growth. However, as determined by Knutsen, making any kind of connection between regime type and GDP size or growth is extremely difficult, since there are so many factors at play. Furthermore, to make the argument that citizens are content with a government as long as there is a large GDP is extremely naïve, and it fails to take into account the larger scope of study, which is comparative politics.

Through my research, it is clear that the determinants of defence spending rely upon far more than regime type and economic growth. The one clear relationship I found was with respect to autocratic regimes spending a high percentage of their GDP on defence. This is a result that leads to the conclusion that autocracies value their militaries highly in order to repress internal strife and to deter external interference. However, no such clear-cut relationship exists in relation to democratic regime defence spending. Although, as expected, democratic regimes spent less as a percentage of their GDP, there was no relationship between democratic regimes and large overall spending, nor between democratic regimes and high GDP growth. In fact, autocratic regimes were found to have a fairly strong correlation with GDP growth. However, there are far too many outliers and external factors to deem this relationship anything more than a statistical anomaly. The largest limitation in my data was the lack of defence spending and GDP growth data for all regimes, which undoubtedly affected my results.

There are many directions that research on this topic could go in the future. In specific relation to this research, it would be interesting to replace the variable of GDP growth with GDP size or GDP per capita. It would also be interesting to examine individual states and their unique determinates to defence spending. Oman for example, ranked highest on percentage of GDP spent on defence at 11.3%, and highest in percentage of total government expenditure spent on defence at an astounding 27.4%. Other states, such as Saudi Arabia, Israel, and Singapore, also could provide interesting examples. Further regional analysis in areas such as the Middle East and South-East Asia would also be interesting, since these areas have seen huge spikes in defence spending since the year 2000. Similar to the study done by Dritsakis, the relationship between GDP growth and defence spending could be examined over a longer period of time to determine if defence spending creates economic growth through the military-industrial complex. However a study of such nature would have to take a more economics-based approach, rather than a comparative politics basis.

In terms of defence spending, Canada, unsurprisingly, remains hidden in the larger “pack” of states, including many of our allies, such as Spain and Australia. Overall, Canadian defence spending has been slowly increasing from a low of $14 billion in 1997, to a recent high of $22.9 billion in 2009. However, in percentage of GDP spent on defence, Canada has been consistently decreasing steadily, and is now approximately half of what it was in 1990. The 2013 percentage of GDP allotted to the Canadian military totaled 1.01%, half the NATO recommendation of 2% GDP. The global increases in defence spending combined with this decrease prompts a number of questions as to how Canada will be prepared for the future, and how we will be able to interact with our allies in the future.

DND photo VL08-2015-0006-007 Sergeant Marc-André Gaudreault

A CH-147F Chinook helicopter in the Base Valcartier training area, 20 November 2015.

Conclusion

Although my hypothesis was not proven to be true, I do not believe it has any implications with respect to the credibility of the Kantian Triangle of Peace. In my results regarding percentage of GDP spent on defence, the Kant’s Triangle was supported, since it was clear democracies placed a lower importance on the military in favour of other spending areas. I believe this finding is far more critical to supporting Kant’s Triangle than the lack of results in my hypothesis are to discrediting it. As seen through the SIPRI annual report, it is clear that defence spending is rising worldwide. Understanding why and how states spend on defence will increasingly be a crucial field of study as long as this trend continues. Although states may look to buy AK-47s, tanks, and aircraft, perhaps understanding why states feel the need to invest in these items could lead to a better understanding on how to reduce worldwide defence spending. And in doing so, we may be able to avoid a situation in which the 75 million AK-47s in the world become 750 million AK-47s.

Oleg Popov/Reuters RTXGA7T

A Chechen fighter armed with an AK-47 stands in a street during an artillery and rocket attack in the centre of the Chechen capital.

Notes

  1. Larry Kahaner, AK-47: The Weapon That Changed the Face of War, 1st edition. (Hoboken, NJ: Wiley, John & Sons, 2006).
  2. Jennifer Brauner, “Military Spending and Democracy”, Defence and Peace Economics 26, No. 4 (October 8, 2014), pp 409–423; Benjamin O. Fordham and Thomas C. Walker, “Kantian Liberalism, Regime Type, and Military Resource Allocation: Do Democracies Spend Less?”in International Studies Quarterly 49, No. 1 (March 2005), pp. 141–157; Julide Yildirim and Selami Sezgin, “Democracy and Military Expenditure: A Cross-Country Evidence,”in Transition Studies Review 12, No. 1 (2005), pp. 93–100.
  3. William Nordhaus, John R. Oneal, and Bruce Russett, “The Effects of the International Security Environment on National Military Expenditures: A Multicountry Study,” in International Organization 66, No. 03 (July 2012), pp. 491–513; Ethan B. Kapstein, “Two Dismal Sciences Are Better Than One—Economics and the Study of National Security: A Review Essay,” in International Security 27, No. 3 (January 2003), pp. 158–187; Fordham and Walker, pp. 141-157.
  4. Spencer R Weart, Never at War: Why Democracies Will Not Fight One Another (New Haven, CT: Yale University Press, 1998), p. 3.
  5. Fordham and Walker, p. 141.
  6. Kapstein, p.175.
  7. Weart, p. 3.
  8. Michael W. Doyle, “Liberalism and World Politics,” in  American Political Science Review 80, No. 04 (December 1986), pp. 1151–1169; Weart, p. 3.
  9. S. A. Bremer, “Dangerous Dyads: Conditions Affecting the Likelihood of Interstate War, 1816-1965,” in Journal of Conflict Resolution 36, No. 2 (June 1, 1992); pp. 309–341; Robert L. Ivie, “Democratizing for Peace,” in Rhetoric & Public Affairs 4, No. 2 (2001), pp. 309–322.
  10. Nordhaus, Oneal and Russett, pp. 491–513; Fordam and Walker, pp. 141-157; Yildirim and Sezgin pp. 93–100.
  11. Nordhaus, Oneal and Russett, pp. 491–513.
  12. J Carter and G Palmer, “What Goes Up Must Come Unevenly Down: Regime Type, Interstate War, and the Dynamics of Defence Spending,” 2008.
  13. Brauner, p. 420.
  14. Adam Przeworski andFernando Limongi, “Political Regimes and Economic Growth,” in Journal of Economic Development 7, No 4 (Summer, 1993), p. 64
  15. Chin-en Wu, “Regime Types, Structural Factors and Economic Performance,” University of Michigan Ph.D dissertation, 2004.
  16. Carl Knutsen, “Political Regime Types and Economic Growth,” 2006. 
  17. Eric Weede, “The Impact of Democracy on Economic Growth: Some Evidence from Cross-National Analysis,” in Kyklos: International Review for Social Sciences 36, No. 1 (February 1993), pp. 21–39.
  18. Kathryn Stoner, “Regime Type and Economic Development,” accessed 26 March 2015, at http://fsi.stanford.edu/research/regime_type_and_economic_development.
  19. N. Dritsakis, “Defence Spending and Economic Growth: An Empirical Investigation for Greece and Turkey,” in Journal of Policy Modeling 26, No. 2 (February 2004), pp. 249–264.
  20. Kapstein, pp. 158–187.
  21. Freedom House, “Annual Freedom in the World Country Scores: 1972-2014,” 2015, at https://freedomhouse.org/report/freedom-world/freedom-world-2015#.VipN5BCrR0w
  22. Stockholm International Peace Research Institute, “SIPRI’s Databases — www.sipri.org,” accessed 22 January 2015, at http://www.sipri.org/databases
  23. World Bank, “GDP Growth by Year”, 2014, at http://data.worldbank.org.