K-Waves and Rebellions – Zach Coles – Medium
A Quantitative Study of the Relationship between a Long-Run Cyclical Economic Phenomenon and Intrastate Warfare
By Zachary Boyd Coles
Under the Supervision of Professor William Belding
Submitted to the Faculty of the School of International Service as a Capstone for the Degree of Bachelor of Arts in International Studies at the American University School of International Service
This study integrates long-run macroeconomic cycles with intrastate state conflict studies, a predominantly neglected topic. Using Nikolai Kondratiev’s theory of a long-run macroeconomic cycle as a starting point, we review other proposed theories of long-run economic cycles and develop a comprehensive hypothesis with the goal of identifying potential new relationships between dynamic economic factors and intrastate conflict.
The hypothesis suggests that regularly timed crises, due to structural changes in the composition of factor payments that are brought about by process innovations, will usher in periods of instability within a state. These periods of instability contribute to the probability that violence will break out within that state. Adopting the Correlates of War dataset to identify conflict and a Kondratiev-Perez hybrid long-run economic cycle to delineate crisis periods, we analyze the years between 1812 to 2007 to test for a statistically significant increase in intrastate conflict during and around crisis periods. Our analysis illustrates a mixed set of results. A regression was ran that failed to show any definitive correlation. However, some scholars propose arguments that support the removal of all datapoints after the fourth oscillatory period in our model. This creates a statistically significant, low-level correlation between the crests of Kondratiev’s oscillatory pattern and the rate at which intrastate violence is initiated.
With the goal of informing future scholarly research, the findings are then discussed within the context of alternative theories and with the recognition of empirical constraints. Finally, we propose further research and implications for intrastate violence studies going forward.
With the growing prominence of insurgent movements since the end of the Cold War, the correlates of intrastate conflict have sprouted into an increasingly important subject. Among other methods, econometric analysis of such disputes has become a popular technique to extract factors that make these conflicts more or less likely to occur. Certain variables, such as GDP per capita or the ethnolinguistic fractionalization index, have been found to correlate with a statistically significant increase in intrastate violence (Elbadawi and Sambanis, 2002). We recognize these variables as indicators. This leads us to question what are the actual reasons these variables correlate with an increase in intrastate violence? Additionally, what causes have we yet to uncover? The reader will find three components within this introduction. First, a brief contextualization of the study is provided, to be expanded upon in the literature review. This contextualization will cover the justifications for the study, as well as the work that informed it. Next, we restate the research problem, the reader now equipped with a better understanding of the forces at play. Finally, we propose a response to the research problem as to aid the reader in understanding how the question has been addressed.
In 1932, a Soviet economist was tasked by his government to argue why capitalism was an inferior system to marxism. Nikolai Kondratiev, by then a well-respected economist and statistician in the USSR and abroad, was given the mandate. He was blindfolded and shot in the back of the head in a labor camp eight years later, having published several studies that reflected the realities of the two systems and how they compared. Clearly, the Soviet government was less than pleased with the conclusions he reached.
Kondratiev analyzed variations in price over an extended period and determined that capitalist societies experienced periods of growth and then subsequent downturns over roughly a five-decade interval. He did not, however, pass a value judgment on whether this was an inferior alternative to the marxist system deeply entrenched in his country. Scholars took up his work and expanded upon it. Eventually, the theory of the Kondratiev Wave was established. Scholars confirmed what Kondratiev had initially theorized: the price oscillations were a reflection of a more complicated story in economics and related fields. Oscillations in price were claimed to be the manifestation of an endogenously generated cycle that emanated from the effect that process innovations have on the relationship between labor and capital holders. While the causal mechanisms behind the phenomenon were criticized and its endogeneity challenged, the Kondratiev Wave’s existence has been empirically demonstrated (Korotayev and Sergey, 2010). Like any wave, both upswing periods and downswing periods exist and comprise the structure of the Kondratiev Wave. Furthermore, its anatomy, within the context of social cleavage and technological innovation, is defined by the differing proportions of factor payments to labor and capital. Perez, a leading scholar of long-run cycles, explained these cycles could be decomposed into four periods with a crisis event between the second and third period.
While the ambiguous language of a ‘crisis event’ does not necessitate intrastate violence, we argue that the decision to rebel against a central authority is function of cost-benefit analysis and that crisis events are capable of providing a catalyst to update those analyses. This blunt perspective is informed by Carl von Clausewitz theory of war and supplemented by many contemporary authors that study intrastate conflict such as Fearon and Laitin. Additionally, because of who these crisis events specifically motivate, solid organizational network are predisposed to arise. Pre-integrated networks and altered perceptions of costs and benefits combine to allow for a higher likelihood of intrastate violence to occur.
Contextualized by the combination of basic intrastate warfare principles, Kondratiev’s initial observations, and Perez’s expansion of Kondratiev’s observations into a coherent theory, a question is developed: do these crisis events indicate that conditions are ripe for intrastate conflict and, separately, does it materialize? These questions, while related, may not be entirely linked. Intrastate violence can appear when conditions are not favorable. Conversely, a conflict may break out at an unlikely time. Perhaps this phenomenon helps explains why intrastate conflict breaks out in less than favorable conditions only. This study, therefore, is specifically designed to address these very questions.
We designed and implemented a quantitative study that examines the number of intrastate conflicts initiated over a period of roughly 200 years that take place, respectively, adjacent to and distant from crisis periods. The rejection of the null hypothesis was supported a statistically significant difference in intrastate conflict initiations surrounding crisis periods when compared to the years outside crisis periods. This analysis is built around a methodology that assumes a reasonable level of rationality and information are awarded to individuals. Additionally, the design adds to other empirically illustrated correlates of intrastate conflict in similar method and rationale. Recognizing that intrastate violence research has been pushed along a certain direction with significant inertia, this study is a necessary divergence from traditional thinking. Although a fairly underdeveloped research area and of somewhat unorthodox elements, the possibility of long-run macroeconomic cycles affecting intrastate violence is a valuable area to expand upon the research that has already been conducted.
Scholars have studied economics and conflict for centuries. Documented instances of intrastate conflict emerged as early as the biblical era. Likewise, as long as markets have existed, people have attempted to dissect them. Enduring in necessity and continuously well-studied, literature about these two perennial topics is abundant. However, one might discover the knowledge base to be comparatively sparse when searching for research that concerns interactions between cyclical economics and intrastate conflict. Although less saturated, groundwork exists to inform how to investigate such an interaction. Those who study intrastate conflict have noted certain environmental, political, and economics conditions under which the materialization of intrastate violence becomes more likely. Economists, especially in the last two centuries, have uncovered cyclical economic events and then theorized their causes and effects as well. We probe a specific question that fits neatly into the niche area of study described above: how the Kondratiev Wave may potentially interact with the rate of intrastate conflict initiation. The lack of overlap requires this literature review to break new ground and discover shared themes. However, with a decent interdisciplinary foundation established, this gap is easily narrowed.
To properly frame a comprehensive theory of how Kondratiev Waves might interact with intrastate conflict, we will separate this literature review into two sections, each of which discuss their respective subjects independently. The intrastate conflict section will emphasize central tenets of logic surrounding conflict and then move to discuss the particularly relevant stipulations of asymmetric conflict. The cyclical economics section, however, will focus on the writings of several key thinkers. We analyze the past work on cyclical economics before moving on to Kondratiev’s theory of the famed “Long Wave’, also called the Kondratiev Wave, and how his ideas interacted with later scholars. In particular, Carlota Perez, an economist at the London School for Economics, most significantly expands upon Kondratiev’s original theory and offers the primary linkage to conflict. Informed by previous scholars’ contributions, we contextualize these complicated interactions and apply them to the current study. Hopefully, it will articulate these ideas in a way that provides a resolute understanding.
CYCLICAL ECONOMICS, THE KONDRATIEV WAVE, AND PEREZ’S TECHNOECONOMIC SUBSYSTEMS
Scholars have long considered that economic conditions may vary in a repetitive, endogenously drive manner. However, until the scientific revolution and fortunate advent of extensive recordkeeping, most studies were inadequate or lost. A shift then occurred several centuries later that allowed for scientists to begin properly analyzing economic phenomena to share with the intertemporal intellectual community. We now have a basket of over 60 different theoretical patterns that scholars have identified and supported with data. We will examine one of the most widely discussed patterns, the Kondratiev Wave, before moving on to later academic influences and expansions. In a separate section, we will discuss the most integral contribution to Kondratiev’s initial model: the inclusion and elaboration of what Perez termed to be ‘technoeconomic subsystems’.
In 1932, Nikolai Kondratiev published his study that outlined the possibility of a long-run price oscillation. His original analysis used historical prices of pig iron and grain in France and England to measure fluctuations in price. These price fluctuations were meant to illustrate the overall level of growth in economies. Kondratiev observed regular intervals of price variability that repeated roughly every five decades. The Kondratiev Wave oscillates because of the cyclical properties of investment and prices. Kondratiev decided his wave was solely composed of and driven by economic factors, as it was not until later that additional subjects became connected to his theory. As the supply of capital in a market increases, the return on investment will typically move inversely in the long run. Likewise, as the supply decreases, return on investment again will rise and incentivize the continuation of the cycle (Kondratieff and Stolper, 1935). Overall, this is a considerable reduction of his theory, but it still provides enough context to understand how later scholars have interacted with and modified Kondratiev’s original idea.
Kondratiev published his studies without truly connecting them to the topic of this paper, but he aligned the building blocks of what became a complicated theory of an economic cycle. Kondratiev’s ideological leanings may trouble some, as marxist economic philosophy has a poor record of success. It is important, however, to note that ideological leanings do not disqualify a claim. It is unlikely his findings were meaningfully biased, considering his politically motivated murder.
Kondratiev reasoned that the price oscillations appeared in five-decade intervals because of technological innovation clustering that occurred on the same timescale. Though claiming the driving force behind the cycle to be innovation, Kondratiev reasoned his model to be primarily a function of the behavior of capital holders. The owners of capital went through periods of prudence that alternated with periods of immoderate investment. He explained later that he believed capital stocks and technological innovations interacted with each other to drive the cycle because a boom in technological innovation would provide new opportunities for capitalists and create a fury of potential profit (1935). The boom would dissipate as market saturation progressed. The economy would then enter a state of decreasing growth, due to the contraction of capital supplies. The faltering economy, he theorized, would then provide an incentive for individuals to experiment with new methods and products to earn their livelihood. The cycle would then repeat with a new explosion of innovation. Although critics argue that Kondratiev did not provide sufficient evidence for this connection, his overall observation of price fluctuations has seldom been challenged. Controversy surrounds the validity of his findings when compared to contemporary years; some economists have found his results to be easily applicable with little or no modification, while others reject the present applicability because of contradictory recent data.
Support for the Kondratiev Wave comes from a variety of scholars. de Groot and Franses synthesized a list of 60+ proposed cyclical patterns and discovered four clusters (2012). These clusters are marked at 8.4 years 21.0 years, 31.6 years, and 54.7 years. 54.7 years, the last number fitting quite well with Kondratiev’s suggested intervals. Likewise, Clement Juglar supported Kondratiev’s idea that macroeconomic cycles were (at least partially) not reflective of exogenous shocks to the world economy. He perceived macroeconomic cycles slightly differently, focusing on an interval of roughly one decade, as opposed to five. His cycle was dependent on fixed investment fluctuations, but he still claimed endogenous generation. A cycle of about two decades was also proposed by Simon Kuznets and was based on demographic swings and their accompanied changes in investment demand. Kuznets considered his cycle to be endogenously generated as well because of the feedback that investment provides to demographic dynamics (2004). The credibility of Kondratiev’s argument is enhanced by these claims, especially the notion that capital stocks drive these cycles.
The claim of endogenous generation is certainly not too far-fetched after considering that independent analyses of world GDP have verified the existence of the Juglar and Kuznets cycles (Koratayev and Tsirel, 2010). de Greene later relabeled the Kondratiev Wave the “Kondratiev Cycle Structure”, being one of the first to imply that there was an interaction between economics and other social forces. Via econometric analysis, de Greene attributed over one third of all price variance in the United States, outside of the secular trend, over last 200 years to four rhythms: Juglar, Kuznets, generational, and Kondratiev. In addition to providing legitimacy, de Greene also created a complication to Kondratiev’s clean model. He argued that the cycle may vary in length (2006). This complication mandates discussion in the literature review because of how it complicates placement of crisis events in our analytical model. Like other scholars, he also elaborated on the structure of a Kondratiev Wave by identifying four periods within each complete oscillation: recovery, prosperity, recession, and depression. Recovery and prosperity are considered to be stages on the upswing of the oscillation, while recession and depression inhabit the downswing. These terms may seem rather defining, but it is important to understand what de Greene was trying to convey. He did not claim that there are long-run oscillations in price that result in steady periods of recovery, prosperity, recession, and then depression. It is not as simple as positive growth and negative growth. Rather, the conversation (here and pervasively throughout the rest of the research) is centered on the idea that, beyond the definite, secular, and positive growth rate of the world economy, there have been oscillations in the overall growth that reflect higher or lower levels than what should have been the average at the time.
Establishing the roots, theoretical underpinnings, and debate around the Kondratiev Wave allows us to consider Carlota Perez’s expansion of the cycle. Her expansion classifies the economic components as only half of the picture, with each phase of the cycle having complementing modules in economic and sociopolitical arenas. Perez approaches Kondratiev’s claim with the shared notion of innovation driving these endogenous oscillations. This perspective draws heavily from Regulation School economics, an academic tradition that traces its roots back to the stagflation that occurred globally in the 1970s. The Regulation School emphasizes regimes of accumulation and modes of regulation. These two concepts can be expressed as such: The regime of accumulation is how capital is moved around and stored within an economy. A mode of regulation is anything that interacts, for better or for worse, with this system of capital movement and storage. The mode of regulation does not have to be strictly economic factors. Perez argues ‘technoeconomic subsystems’ are the result of modes of regulation and regimes of accumulation interacting. A technoeconomic subsystem that arises due to grand technological process innovations is fundamentally different than the subsystem that existed directly before. The socioeconomic institutions encompassing the agreed way to move and distribute capital are no longer set in harmony with the new possibilities of capital accumulation allowed by these process innovations (Perez, 1983). Growth does not stop. In fact, growth accelerates. Perez divides the subsequent interaction into five different periods:
1. Irruption: First, there is an irruption to the system when an old method of production becomes obsolete in the face of an improved method. Following the trough of a wave, the opportunity for innovation to create growth is at its peak level. A process innovation, something that fundamentally alters production, an invention like the steam engine, is developed. During this early upswing, there is explosive growth and innovations quickly expand upon the new technological paradigm. Many old industries become obsolete during this era, accompanied by a rise in unemployment.
2. Frenzy: A frenzy of investment then occurs midway up to the crest of a wave. Owners of capital seek to invest in this new method and its accompanying industries. Industries begin consolidating gains and completing all possible capitalization of the new technological paradigm. This ends at the crest of the wave, as the last new products are developed and the market reaches relative maturity. Perez notes this segment will show heavy investment in the process innovation technology and a notable separation of the rate of return between the rich and poor.
3. Crisis: A crisis event occurs. Perez leaves this largely undefined. After the crisis event, the mode of regulation is harmonized with the regime of accumulation. The technoeconomic subsystem has returned to a stable position
4. Synergy: Following the readjustment, growth and production level off with realigned and more equitable factor payments between labor and capital. This segment of the wave can be characterized by hope and full employment. Individuals experience the ‘most fair’ version of capitalism in this era. There is synergy in the system as all parties begin to experience the benefits of the more productive technoeconomic subsystem.
5. Deployment/Maturity: Lastly, proper market saturation occurs as the new system reaches maturity with its dual components now in line. The economic paradigm that the process innovation brought about decades earlier has now fully developed. Growth levels off in this era. Individuals are still reaping the benefits of the more equitable distribution of gains. However, the climate is now favorable for ideological skepticism, as many begin to see the unfilled promises of a ‘Great society’ made in response to the crisis event become unfulfilled.
Because of the decreasing factor payments to labor as the second period progresses, owners of labor lose the incentive to remain within the confines of the existing socioeconomic agreement that was previously forged and codified by the establishment. The crisis can be resolved if owners of capital and the state agree to alter the socioeconomic agreement. If not, then the owners of labor can, in theory, react violently in order to alter the agreement to return to more equitable factor payments results. This is the basis of intrastate conflict occurring within long-run macroeconomic cycles.
The foundations of intrastate conflict scholarship are derived from centuries-old tacticians and contemporary scholars alike. It is a unique academic space in which the conversation extends millennia. Along with pure conflict research, tangential areas of importance include decision-making studies (also sometimes considered to be an extension of economics) and the subsection of sociology that specifically analyzes the dynamics of group recruitment and activity. With these topics in mind, this section of the literature review will explore questions that relate to the reasoning behind conflict initiation, network effects, and when conflict is most likely.
Lastly, this paper addresses intrastate conflict, not interstate conflict. Thus, a different definition is provided. Intrastate conflict occurs between a central government force and a non-governmental force. As many theoreticians and philosophers have argued, it is imperative for the state to fulfill its role as an entity that maintains a monopoly on violence to be used as the ultimate arbitrating force in disputes and to provide protection against external actors. Therefore, an intrastate conflict has traditionally been overwhelmingly asymmetric. A strong, central power is challenged by a force with fewer resources and organization.
War against a stronger opponent is generally ill-considered. The chances of success have proven to be poor. An analysis of insurgencies in the last 40 years by the Rand Corporation illustrated that, at best, an insurgency is less than 35% likely to achieve victory within 32 years. After that, the rate of success drops to closer to 0%. The median length of an insurgency’s survival is about ten years, assuming they survive through the proto-insurgency stage in which networks are developed (Connable and Libicki, 2010). Accounting for the grave outlook reflected in those statistics, what led to the decision for civilians in Syria to challenge the authority of the Assad regime in 2011? There are factors that push the unfavorable chances of success to be preferable when compared to the alternative of continuing to recognize the regime as the legitimate political authority. Faced with a competent foe, no rational actor would willingly limit their capacity for organization, military strength, and logistics that were available. It would be foolish and dangerous. Applying a rational choice lens, the chosen lens for this study, we must assume that these actors either operate with egregiously incomplete information or that they consider the status quo to be utterly intolerable. Scholars argue for some of both. In any case, we see that both arguments strengthen the case for interaction with the Kondratiev Wave. Slantchev introduces a model that draws from Carl von Clausewitz’s bargaining framework that describes conflict simply as the failure of two parties to reach an agreement. He extends this analysis to asymmetric conflict, stating that the outbreak of rebellion is the tipping point in which the status quo of state control becomes intolerable. It is at this point that the dangers of losing a war against that state appear less severe than the drawbacks of continuing to live under the state. The reader is cautioned not to interpret this claim too literally. It is not the case that the state does something to render the previous status quo obsolete. Instead, Slantchev is arguing that the result of many shifting factors (possibly outside the control of the state itself) is an altered perception of the costs of remaining under the state compared to those of rebelling against the state (2003).
When asking why actors choose to initiate asymmetric conflict, we must look to all perspectives and sets of assumptions. Slantchev’s argument is comprised of a dispassionate, objective analysis. However, there is also a less rigid set of assumptions. Instead of a rational, analytical approach to conflict initiation decisions, dissenting opinions argue that because of the personal and emotional nature of intrastate warfare, certain actors are more likely to engage in irrational behavior. McCormick supports this idea by identifying the difference between rationalist and expressionist terrorists, who are actors using a tactic commonly reserved for asymmetric conflict (2006). Rationalists attempt to push change in a system, while expressionists convey their dissatisfaction with the system. The end result may be the same, but the aim of this paper is to examine justification.
The expressionist model is certainly something to consider, but Caplan’s analysis of suicide terrorism arises as a foil. Caplan presents an argument that is based on the practicality of relevant academic research, stating that even when these irrational outliers persist in data, a rational behavior and expectations reference is a valuable starting point for analysis, if not the default assumption. When we propose that intrastate conflict is sparked by people who are no longer satisfied with the status quo, we assume these people to be rational actors that are engaging in conflict with a goal in mind. Thus, the model of rationalist terrorism, — and all other forms of asymmetric conflict — is the primary perspective. Conflict with purpose is the most logical frame. Bryan Caplan supports this argument in his application of the rational choice model to terrorism. He argues that the willingness to act irrationally declines as the costs of irrational action climb. It is reasonable to argue that the decision to engage in violence under irrationality is extremely costly (2006).
Equipped with the model of why actors initiate an asymmetric conflict, we can now put forth a comprehensive theory that helps explain when actors decide to initiate an asymmetric conflict. An actor will commit to waging an asymmetric war against the state when the actor believes they can win due to a superior ability to hurt or endure costs, irreconcilable differences (existential threats), or if they feel they will face existential threats in the future.
We can return to a classical military theoretician’s analysis of what causes war. For instance, Carl Von Clausewitz’s explanation for the outbreak of war can be briefly summarized as such:
“Fighting breaks out when two sides cannot reach a bargain that both prefer to war. Each side fights to improve its chances of getting a desirable settlement of the disputed issue. The war ends when the two sides strike a bargain that both prefer to the continuation of hostilities. The outcome is the bargain that is struck. Finally, the duration of peace following the war reflects the willingness of both sides not to break the war-ending bargain.” (Reiter, 2003)
Several modern versions allow for modification of the framework in order to include more sophisticated interpretations of actor capacity. Thyne offers the distinction between the power to hurt and the power to endure costs. He describes the distinction by noting that a non-state actor often has disproportionate abilities to hurt the state actor, with the most well-known example being terrorism. The general public has a much lower tolerance for bombing casualties than freedom fighters do (2006). Reflecting on the research question, we must consider that, while a non-state actor’s absolute power to hurt is significantly limited, the state has a lower threshold of tolerance. Likewise, the non-state actor has considerable power to endure costs relative to that of the state. Fearon and Laitin summarize and contribute to these points, claiming war is only possible under three different circumstances:
1. Disagreement regarding the outcome of a conflict will promote war. Essentially, imperfect information regarding power to hurt and power to endure causes war.
2. War may occur because of an inability to commit not to fight in the future.
3. Bargaining may not be possible and cause war if the material to be divided is indivisible. (2003)
In the context of Kondratiev’s theory and Perez’s elaboration, one can conclude how these qualifications might be met within crisis event years more than the years outside a crisis event.
We must also acknowledge the role of preexisting networks in facilitating initiation of asymmetric conflict. Staniland claims that pre-conflict networks of rebellion influence how violence takes places and the effectiveness of the organization. He continues to note how organization cohesion may come from a variety of sources, including ethnicity or shared economic background. Staniland allows for the possibility of the organizations to change in structure and composition, but does also note the difficulty of doing so (2014). Depending on the levels of social and vertical integration of the pre-conflict organization, a group will find it either easier or harder to successfully coordinate asymmetric confrontation. If we are to continue with our consideration of rational actors choosing to rebel when the benefits outweigh costs, then the apparent organizational ability of a group will too play a role. Using Staniland’s claim that groups with shared identities and spaces are positioned to more effectively coordinate, we argue conflict initiation is more likely to meet the threshold required to be considered as an instance of intrastate conflict initiation because of the diminished possibility the rebellion will immediately fail.
Asymmetric conflicts in which the non-state actor has a common group composition and previously proven functional capability will promote the survival and success of a non-state actor. Additionally, the commonalities and previous capabilities may help remove the principle-agent issue that plagues many organizations. This is an area where the literature may work against the proposed hypothesis. Knowing that labor power is diminished leading up to crisis events (as holders of labor have fewer resources and say in governance) networks may have less cohesion than other times (Perez, 2014). Alternatively, instances of worker solidarity often occur when facing what is deemed as unfair bargains. We infer, however, this is somewhat culturally bound.
For all of these concepts, the word ‘cause’ is somewhat of a misnomer. No single variable commands non-state actors to commence warfare against the state. Instead, social scientists have commissioned a number of studies in recent decades that strive to prove empirically that some conditions make asymmetric conflict more likely. For instance, the most proven variable (although probably the least relevant to this particular study because of the variables static nature over time) is ‘rough terrain’, identified and defined by Fearon and Laitin to significantly increase the likelihood of non-state actors engaging in violence against the state (2003). The increased likelihood does not imply causality. In fact, this is their central claim. The cause is simply an illustration of weakened state capacity and strengthened non-state actor capacity. This in mind, we can identify a plethora of variables relevant to cyclical economics that would appear to create favorable conditions for asymmetric conflict to break out.
Starting with our exemplar authors, we can identify several factors instrumental in increasing the likelihood of intrastate warfare. For example, they include poverty and political instability. Contrary to other authors, Fearon and Laitin do not claim that poverty is an underlying cause of intrastate conflict, but simply a marker of a financially insolvent and bureaucratically inept state. They qualify political instability to include a newly formed nation, an anocratic regime, or political instability at the center (2003). Since economic cycles create periodic busts, we can infer that a state is more likely to be financially insolvent during certain parts of a cycle. Political instability at the center is relevant due to the increased militancy of labor that is described to be integral in Perez’s explanation of a Kondratiev Wave. More specifically, prior to the occurrence of a crisis event, a large segment of the population is definitely worse off than others who are benefiting from the shift in the regime of accumulation. The authors also find that several variables are not predictors of increased likelihood asymmetric warfare occurring. Income inequality and ethnolinguistic diversity do not seem to indicate an increased likelihood. However, Collier and Hoeffler interpret the insignificant results found by Fearon and Laitin differently, arguing that the motivations of ethnolinguistic tensions and income inequality may be underlying factors, but they can only be catalyzed by certain socioeconomic and political opportunities (2002). Collier and Hoeffler also take issue with the common methods of data collection and interpretation in regards to income inequality. The common measure is the Gini coefficient, which only captures inequality between individuals. They claim that horizontal inequality, inequality between groups, may yield more significant results (2002). Stewart strengthens this argument by presenting his own case for horizontal inequality causing crises (2000).
As we will discuss in this literature review, the foundation of an organization can be fundamental in fomenting action. To misuse an old adage, you go to war with the organization you have. Jazayeri expands this observation by demonstrating that the relationship between intrastate conflict and identity-based inequality, horizontal inequality, is particularly pronounced in the MENA region (2016). These networks are a representation of a disadvantaged demographic, supporting the previous argument that income inequality may allow for greater group cohesion. Buhuag builds a study around the idea of an alternative method of measuring income inequality, showing a statistically significant variation at the p<.01 significance level (2013). Thus, the verdict must remain out on income inequality. Though not crucial to support the overall logic, we can continue allowing this idea to be an auxiliary support to Kondratiev’s and Perez’s model.
Hoeffler continues the conversation in another publication, outlining the necessity for both greed and grievance. He argues that there must be an opportunity, such as a weaker state and enhanced social cohesion of the group that would rise up against it. Likewise, the chance for a better outcome must exist — the greed component. Hoeffler frames rebellion as a public good. “Since rebellion necessitates private gains even if it is based on common grievances, theories of civil war should consider both factors.” (2011). He cites Tullocks’ notion of the “paradox of revolution” (2005), where rebellion is considered to be a public good that benefits everyone if the movement is successful. However, private benefits of individuals who join a revolution are important and become smaller as the group size swells. This implies there is a sweet spot of rebellion in which a smaller group will find it easier to reward members, but a larger group will find a higher probability of success and attainment of that public good. He considers a solution to this issue by examining the methods of recruitment by insurgencies, including some that incentivize joining via artificially raising the opportunity cost of not joining, sometimes labeled as forced conscription (2011). This is often the case with intrastate conflicts. Though highly condemnable, this solution does somewhat negate the problem of obtaining a critical mass of participants. The positive and negative incentives mentioned above, as well as the irrationality of individuals, can all be used to address the free-rider problem of the public good of revolution.
When considering all of these potential causal mechanisms, there should be a lingering question of how a particular variable changes the cost-benefit analysis for a potential rebel. We all have a metaphorical price at which non-state violence becomes favorable to the status quo. Each of these variables have been argued by econometric analysis or by extensive theory to suggest an alteration in the cost-benefit analysis will occur. Crisis events that take place within the crests of Kondratiev Waves may bring these components together and increase the likelihood of intrastate conflict.
These idea have already been demonstrated. For example, Peter Turchin examined political instability in the United States for its entire existence through a demographic lens. He showed that there was a strong pattern of bigenerational violence that have periods of slightly more than 50 years, roughly lining up with Kondratiev’s proposed cycles. Turchin offers an explanation grounded in structural demographic theory, which claims that a surplus of labor leads to decreasing standards of living that cause a wave of prolonged and intense sociopolitical instability. Turchin’s analysis intersects with a description of Kondratiev Wave upswing where labor loses bargaining power and Hoeffler’s necessity of a lowered opportunity cost of forgoing rebellion. Labor is employed less, paid less, and experiences diminished bargaining power. As all this occurs, the relative cost of joining a rebellion decreases and the chances of finding an alternative that is more profitable declines as well (2012).
Lastly, the assumption that intrastate violence must be asymmetric is addressed. If this were not the de facto rule, then the variables affecting violence initiation would be different. Kalyvas and Balcells support the notion that an asymmetric warfare strategy is an inevitability and not a preference. However, in the second part of their argument, they makes a bold claim. They argues asymmetric warfare was simply an artifact of necessity that non-state actors were forced to adhere to because of the geopolitical restrictions imposed by a bipolar world order existing during the Cold War, in which one of the two superpowers would artificially prop up a weakened state. Non-state actors could not move on past asymmetric warfare to directly confront the state because of the support stemming from one of the two superpowers. Kalyvas and Balcells claim that, with the fall of the USSR, the incentive for the remaining superpower to prop up weakened and unpopular regimes has diminished. Thus, non-state actors should be able to pass through the asymmetric warfare stage and move towards what he calls symmetric non-conventional conflict (both actors have low capacity) or conventional conflict (both actors have high capacity). Essentially, asymmetry is eliminated (2010). However, their assumption that states will be unable to sustain their previous capacity may run into some issues. Capacity does not erode quickly after it has been established; an armored battalion does not go out of service very quickly. Admittedly, the resources required for upkeep may dry up, but it is important to note this caveat. The incentive for a geopolitical superpower may be gone, but regional powers still maintain the same ideas and act on them accordingly; though the competition between two superpowers no longer exists, regional competition is just as prevalent. Their argument also ignores centuries of precedent and contemporary evidence that all suggest asymmetric conflict is not necessarily contained within the epoch of a bipolar world order. Undoubtedly, there is utility in external support. The absence of external support has been proven to reduce the likelihood of asymmetric conflict success, but not necessarily the outbreak. Some may challenge this assertion because we stated individual cost-benefit calculations (where external support would be taken into account) to be integral in the decision to initiate intrastate violence. However, all of these contributing variables are in amorphous and relative terms. We presume this will have a negative effect on the propensity for an individual to engage in rebellion, but have no reason to consider the magnitude of the effect to be more significant than each of the other factors pushing the opposite direction.
This literature review involves conceptualization of possible interactions between complicated forces such as the cyclical nature of capital investment, innovation clusters, rebellion social network composition, bargaining frameworks, and violence as a public good. This is not holistic; scholars could (and have) written entire anthologies on just one of these topics. Rather, it is an overview to inform the methodological justifications and analysis of the research question. In regards to cyclical economics, theories of cycles that detail the role and effect of various levels of capital accumulation and labor capacity are at the forefront. The main criticisms of these cycles are the ongoing debates about whether the phenomena are endogenously generated, a fundamental part of classifying something as a wave; and the lack of a clearly defined causal mechanism. These epistemological shortcomings were fundamental in developing the proper method of preparing and analyzing data. Using a rational actor lens, the justification for intrastate conflict was defined within a modified version of Carl von Clausewitz’s bargaining framework. We also determined that intrastate conflict still is mostly contained to be expressed as asymmetric conflict. An actor will initiate violence against the state only after that threshold of costs has been cleared, providing an answer to when the conflict starts.
Methodology and Methods
RESEARCH PROBLEM AND CONTEXTUALIZATION
Nikolai Kondratiev argues that a macroeconomic phenomenon occurs at approximately 50-year intervals. According to Kondratiev and a cadre of later academics, the cycle’s mechanics include bigenerational disruptions to a technoeconomic subsystem due to technological process innovations that alter the mode of capital accumulation. The previous social framework struck a balance in bargaining power between those who possessed labor and those who possessed capital. As a new technoeconomic subsystem emerged, the balance became misaligned with the new mode of capital accumulation favoring those who possess capital. Altering the opportunity costs of participating in the society, the misaligned social order provides the necessary motivator to abandon the existing social framework. It is possible (and perhaps likely) that this will not be accomplished peacefully. Existing literature points to a possible correlation existing between the Kondratiev Waves and intrastate conflict and provides the underlying necessitation for study. We should and must endeavor to investigate intrastate conflict, always improving our explanatory and predictive powers. The power to predict intrastate conflict is especially important when considering that the implications for civilians in intrastate conflicts are often far more grave than for those of interstate conflict because of the enhanced utility of humans as a resource for non-state military forces and the blurred distinction between combatants and non-combatants. A better anticipatory posture that has been informed by enhanced predictive capabilities can save innocent lives.
The study’s interdisciplinary nature allows for several social science frameworks to interact with each other in fairly unique ways. These frameworks create a perspective to evaluate and test arguments, based on a network of assumptions made. The interaction extends across the separate fields of economics and conflict studies. Common linkages between the two areas used extensively include the omnipresent principles of rational choice and opportunity cost. Additionally, the neoclassical traditions of economics are accepted with no more than negligible challenges to the standard assumptions and paradigms.
DESCRIPTION OF METHODOLOGY
The null hypothesis proposed is that Kondratiev Waves crests are not associated with more intrastate conflict than the rest of Kondratiev Waves. In this case, Kondratiev Waves act as the independent variable interacting with dependent variable of intrastate conflict.
Null Hypothesis: The initiation rate of intrastate violence during the crest of Kondratiev Waves is equal to that of the rest Kondratiev Waves
Alternative Hypothesis: The initiation rate of intrastate violence during the crest of Kondratiev Waves is greater than that of the other parts Kondratiev Waves
If the null hypothesis is rejected, then we expect there to be a statistically significant, positive, low level correlation with intrastate conflict and Kondratiev Wave crests. We suspect the results could become less statistically significant as time progresses because of the mutable intervals which could be decreasing in length, as some scholars have argued (Narkus, 2012). We expect the lowest level of significance to occur in the interval that immediately follows the conclusion of Second World War due to irremovable confounding variables, such as the rise of a bipolar world order heavily focused on indirect state competition that encouraged low-intensity conflict and proxy wars.
The respected Correlates of War (CoW) project, an ongoing effort to catalog organized violence that extends from 1817 to 2007, operates as the dependent variable. The subcategory of data that is housed within the intrastate conflict dataset, was small enough to input the entire population into the regression. The other subcategories of data are housed within interstate conflict dataset, non-state conflict dataset, and the extra-state conflict dataset. These subcategories were all excluded from the analysis. The interstate conflict dataset was excluded for obvious reasons and the extra-state dataset, while sometimes including instances of intrastate conflict, generally was an account of foreign expeditions by states and then subsequent attempts to repel the aggressor by domestic entities. These conflicts, while sometimes coined to be rebellions, are more a reflection of extraterritorial aggression than internal strife. The non-state dataset includes conflict that did not involve any central authority, an integral actor for a conflict to be considered intrastate. While O’Neill notes the near-ubiquitous nature of external support in most intrastate conflicts and certainly any successful intrastate conflict, the purpose of this study is to examine intrastate conflict generated endogenously and internally, rather than a reaction to expelling a foreign power exemplified by conflicts like the Afghan resistance to NATO occupation (2006). CoW datasets were derived from a variety of authors over several decades, originally created by David Singer and Mel Small to develop a typology that could differentiate between various types of war. The database has been frequently updated since its original inception and underwent a distinct methodological alteration at the turn of the century that redefined how intrastate and extrastate wars were classified. The result was an expansion of what could be considered an intrastate conflict and a more realistic method to determine qualifications surrounding combatant casualties.
The CoW project has defined intrastate conflict as events that fulfill the following parameters:
1. Total deaths in the conflict must exceed a threshold of 1,000. Total deaths include state deaths, non-state actor deaths, and civilian deaths. Civilian deaths were justified as includable for two reasons. First, the line between combatant and civilian is often blurred in intrastate conflict, making it difficult for those collecting data. Second, due to the nature of intrastate conflict, civilian fatalities are often intentional. The non-state actor can be considered a war participant if it either commits 100 armed personnel to the war or suffers 25 battle-related deaths. A state is considered a war participant if it commits at least 1,000 troops or suffers 100 battle-related deaths. Note the differentiation between total deaths and battle-related deaths.
2. There must be an “effective resistance” by non-state actors. This can be defined two ways:
a. Both sides must initially be organized for violent conflict and prepared to resist the attacks of their antagonists
b. The weaker side, although initially unprepared, inflicts upon the stronger opponents at least five percent of the number of fatalities that it sustains. This is meant to differentiate wars from massacres or riots
The conflict must take place between the de facto governmental authority, often termed to be the state actor, and one or more non-state actors. If an external intervener enters the conflict, then it remains an intrastate conflict, but becomes ‘internationalized’. If the external intervener enters the war and takes over the bulk of the fighting for a non-state actor, then it becomes an interstate war. Lastly, if the external intervener enters the war and takes over the bulk of the fighting for the de facto government, then it becomes an extra-state war. In an attempt to purify the dataset, the intrastate conflict dataset was edited to remove instances of a conflict appearing more than once due to a number of instances in which specific actors entered or exited a conflict. The main piece of information that must be pulled from each data point is the year in which the conflict began. No other alterations were made to the data that comprises the dependent variable. Both a blessing and a curse of this project is that the population remains small enough to analyze the entire record of incidents, rather than a sample of the population. However, we must recognize the limited amount of data is more likely a result of the short history of comprehensive record-keeping. This is unfortunate and applies to the independent and dependent variables alike. However, as with all social science, the data available, while imperfect, can be used to create a powerful study that determines whether there is a relationship between the variables. Of course, no data collection method is without flaws. The CoW database suffers from the usual restrictions that many other databases encounter. The CoW is dependent on a variety of sources for each entry and, while completing its due diligence in corroborating all data, is prone to source bias or error. Additionally, this cannot be a controlled experiment in many other fields. Thus, we are reliant upon an imperfect collection of data from a limited period of time.
The independent variable is an attempt to objectively interpret the writings of Nikolai Kondratiev and later scholars to create a timeline in which the crest of each Kondratiev Wave exhibits an increased amount of intrastate conflict.
For the sake of this study, we can approach the disputed predictive power of a Kondratiev wave in two ways. First, we can accept the claims made by previous authors that the Kondratiev Wave is an empirically demonstrated phenomenon, despite reservations. Alternatively, we can hold off on accepting the empiric nature of the Kondratiev Wave, but still run a study and apply a ‘cross that bridge when we get there’ style of approach. Either approach yields limited implications for the study or the supporting methodology. If there is no correlation, then it did not matter. If there is a correlation, then something is creating 50-year cycles of intrastate violence. The latter situation would warrant further discussion just the same. Perez does make a distinction between core and periphery economies, emphasizing that technologies diffuse to the periphery with a lag (1983). However, this does not mean that crises are not relatively synchronous. The reason for not including a periphery-based crisis lag is relatively straightforward. Because the rapid pace that financial markets move, even 200 years ago, financial panics take a relatively short period to diffuse across markets. While the social and political polarization may not have reached its potential in the periphery markets that the financial crises are diffused to, it is reasonable to assume that the fluctuations in finance are enough to set off a similar, if somewhat muted response in the periphery markets.
The overall test design is not overwhelmingly complicated. A regression was developed that was labeled ‘intrastate conflict initiations per year’ as the dependent variable on the Y-Axis. This is simply the numeric value of the number of intrastate conflict initiated in a given year. The X-Axis was labeled ‘distance from crisis event’ as the independent variable. The independent variable is the numerical value of a year minus the numerical value of the nearest crisis year, then turned into an absolute value. Crisis events occur exactly once during a complete oscillation, meaning once every 50 years. For instance, a crisis event exists in the year 1915. If the datapoint is 1921, then it will be placed at a distance of 6. Since the cycles appear in intervals of 50 years, the maximum distance a year can be from a crisis event is 25 years because crisis events exist before and after any year being measured.
Based on our interpretations of arguments made by Kondratiev and Perez, there is no reason to believe that years before the crisis would interact with said crisis point any differently than years after a crisis point. Additionally, it must be kept in mind that the crisis event, a singular year, is a nominal value. It is possibly and likely that actual Kondratiev Wave crests occur several years before or after the value that is given. However, it becomes increasingly less likely that symptoms indicating a crisis event will occur as we move further away from when the crisis event is said to occur. After defining these variables and situation them into a linear regression we arrive at the actual concept by tested. If there is no correlation between Kondratiev Wave crests and intrastate violence, then the slope of this regression should be zero. However, if the null hypothesis is rejected, it will be illustrated by a negative slope as the amount of intrastate conflict per year decreases the further away from a crisis event a datapoint is.
Crisis event date selection was initially problematic. Kondratiev argued for certain years that these Kondratiev Wave crests, Perez’s crisis events, occurred in. However, the academic cohort took issue with some of the dates, slightly deviating from them. In order to mitigate any issues that would arise from errant crisis event date selection, an average of the dates proposed by all major scholars was taken. We determined that the average dates proposed by all scholars followed Kondratiev’s initial claims very well, with the largest deviation being seven years and most deviations much smaller. Thus, Kondratiev’s dates were used in the study.
We place crisis event dates at Kondratiev Wave crests because of Perez’s arguments about how crisis events are created. Crisis events are not the low point of growth. Rather, crisis events are the high point of unsustainable growth. Perez supported her claims by situating crisis events around periods where real growth leveled off (1983). These events demarcate periods of time in which returns are enormous because of the undelivered payout that owners of capital can withhold from society, all due to a misaligned technoeconomic subsystem that is not cohesive with the preexisting social framework.
We pushed for a Large-N quantitative study to most conclusively determine a relationship. Considering there are 190 years of unique intrastate conflict initiation observations, a case study approach is ill-suited for this research question. It is crucial to not perpetuate more subjective interpretation and ambiguity into the already controversial nature of Kondratiev’s work. Approaching the research question from the perspective of measurement and not meaning, we determined that an interpretivist design is improperly suited for this task as well. Thus, a Large-N quantitative analysis is the most logical. Lastly, the decision to use Kondratiev’s own definitions, rather than an indirect measurement of the wave’s properties (such as a Gini coefficient), was necessary to keep with the spirit of determinant findings. No suitable measurement has been developed that extends far enough back in time.
The bird’s eye view aside, there were several frameworks or assumptions incorporated into the procedure and analysis that must be justified. These include highly theoretical questions such as rational choice theory or the proper reconciliation between neoclassical economics and a theory tainted by the reputation of Marxism. The other justifications are answering particular studies or claims made by scholars that are particularly relevant to this study.
The decision to base when an individual enters conflict on a rational choice model is one that is informed by several renowned scholars in the conflict and strategic studies fields, some predating what would be considered modern academia. For instance, famous Prussian strategist Carl Von Clausewitz first outlined a bargaining framework model in his unfinished novel On War. His theory of conflict initiation and termination was founded on the idea that nobody engages in war for the sake of war; all conflict must have a larger goal. Furthermore, the expected benefits of achieving a goal must outweigh the expected costs of engaging in war (2013). The application to intrastate conflict is as follows: before the technoeconomic system became misaligned with an outdated social system, the costs associated with challenging the central authority outweighed the benefits of a new authority. Following the misalignment, the cost-benefit ratio has shifted to include higher benefits. If we are to combine this assumption with a rational choice model, then it would be expected that a rational individual would choose to participate in an intrastate conflict. An individual familiar with the paradigm can easily refute this claim, arguing the average person is not capable of discerning this new information or effectively interpreting it. The rational individual makes choices with the information in their possession, so it could be argued that an individual in the misaligned system would now have information that would lead them to believe the costs of conflict were higher than the benefits of conflict. This is a possibility, but there is nothing to lead us to believe that a systematic bias would misinform individuals as costs and benefits change. Rather, it is much more likely that bias is randomized, affecting individual decisions but not affecting the overall number of participants due to the law of averages. Bryan Caplan, mentioned in the literature review, emphasizes this in his book titled Myth of Rational Voter. Furthermore, he proceeds to claim that, as the costs of being misinformed go up, individuals strive to be more informed. Random and systematic biases are more likely to be eliminated (2008). In the context of intrastate conflict, being misinformed may result in death. The costs of being misinformed are prohibitively high and would encourage individuals to make decently informed decisions. Regardless of the random distribution of risk propensities in a population, a misaligned social system will, on average, raise the benefit of intrastate violence relative to the cost of said violence.
It may seem peculiar to approach a hypothesis forged in the crucible of Marxism with such neoclassical economic assumptions. The reality is that, in order for an empirical analysis that has roots deeply in the profession of economics to be legitimate, it must approach the social science of economics from a neoclassical perspective. However, we largely agree with neoclassical economics without the added pressure of academic acceptance to begin with. Thus, the decision to comply with and accept the notions of neoclassical economics was never in question. Rather, the question was if Kondratiev’s work could fit in a neoclassical analysis and if some rules could be bent for the sake of accepting a conclusion drawn from a Soviet economist. There is a historical precedence here as well. Wassily Leontief, a Soviet economist, famously published work in 1953 that outlined a paradox found in international economics based on the neoclassical Hecksher-Ohlin model of international trade (1953). This is widely accepted and taught today. Historical precedence aside, after considerations based on the source of Kondratiev’s information and his procedure, there was nothing that did not comply with basic economic rationale. His motivation and ideology had decidedly Marxist tinges, but this did not affect the outcome of the study. In fact, Kondratiev’s work was lacking Marxist tendencies to such a degree that the Soviet government saw to it that he was executed. Clearly, his work does not deserve the level of suspicion that the initial glance garners. Additionally, a popular retort to the idea that an extreme amount of capital inequality, a concept fundamental to this study, does not produce strife and instead inspires those at the bottom to work their way up is not appropriate: the social system misalignment allows for less social mobility than before. Whether this drives the whole population to engage in intrastate conflict is beside the point — the result means there will be an absolute increase. An absolute increase, especially when put in the context of rational choice becoming increasingly prevalent as the costs of irrationality increase, will plausibly encourage intrastate violence. Kondratiev may have had dubious motivations, but his study is valid when accounting for the empirical design and specific deviations from neoclassical economics that would be relevant to this study.
Initially, there was the temptation to claim pervasive applicability because of the wide range of time examined and holistic nature of the datasets. That is, unfortunately, not the case. While 400 initiations of conflict over 200 years may seem sufficient, the dataset pales in comparison to the history of intrastate conflict. This is a history that reaches back at least 3000 years. 7% is paltry amount of history to have covered. There is, however, another complicating factor. The above statement is grounded in the assumption that the circumstances facilitating every creation, maintenance, and destruction of an organized political entity are uniform and perpetual. Instead, the mechanisms are largely ephemeral when placed in the context of the larger picture. This is precisely what Kondratiev argues: each 60 year epoch changes the mode of capital accumulation in such a way that the supporting structure of society must adjust to maintain harmony. This is a double-edged sword though. Kondratiev assumed his wave was a result of economies exiting a feudal system (undoubtedly an opinion informed by his ideological leanings) and did not devote any writing to pondering whether or not this cycle would be everlasting (1935). Moreover, scholars who followed Kondratiev and later commented on his work frequently claim that the cycle has theoretical justifications for accelerating and have demonstrated that. Evidence of this claim can be found especially in the fifth interval of the Kondratiev Wave (Narkus, 2012). In short, it is unknown if Kondratiev’s theoretical phenomenon has a shelf life of 250 years or of 2500 years. We must treat the Kondratiev Wave with the knowledge that it is possibly transient event. However, just as there is no reason to assume it is applicable 100 years from now, it is equally as fair to claim there is no reason to assume that it will not be applicable 100 years from now. In a slightly different vein, the applicability across geographic or categorical differences of intrastate conflicts remains relatively strong. If the study incorporates all types of intrastate conflicts across the globe, there is a strong argument for the applicability to hold, if only in the current moment, without limitations of location. Concisely, in terms of weaknesses, this study’s primary flaw is that there is no evidence that Kondratiev Waves will continue to exist in their current form in the future.
Following data aggregation and the appropriate regression construction, several key findings emerged. First, examining the data set in its entirety yields no discernible relationship between intrastate violence initiation and Kondratiev Wave crests. The correlation coefficient is miniscule at -.03. The trendline slopes downward at the diminutive rate of -.007x. The R² value is similarly small at .0007. With these results we cannot conclusively reject the null hypothesis that there is no relationship between Kondratiev Wave crests and intrastate violence. With a P-value of .35, there is 35% chance the results were obtained due to random data variation.
A secondary regression, however, revealed results that were identified to be statistically significant. The secondary regression includes the dataset until the fourth Kondratiev Wave trough that occurs in 1940. The trendline slopes downward at a rate of -.029x and is accompanied by a correlation coefficient of -.15. However, at .025, the R² value produced is still small. These results, as demonstrated by a p-value of .0395, are statistically significant and can be interpreted to mean that there is less than a 4% chance these results were obtained randomly. , the numbers indicate a 3% reduction in intrastate violence initiation for every year further away from a Kondratiev Wave crest and that 15% of intrastate violence initiation variation can be attributed to this trend. In the secondary regression we reject the null hypothesis that there is no relationship between Kondratiev Wave crests and intrastate violence initiation.
The results for the entire time period analyzed do not support a rejection of the null hypothesis. The regression results are shown below.
Nominal Peaks Simple linear regression results:
Dependent Variable: IS Conflict
Independent Variable: Closest Distance (Years)
IS Conflict = 1.8296529–0.0067480927 Closest Distance (Years)
Sample size: 190
R (correlation coefficient) = -0.028061519
R-sq = 0.00078744883
Estimate of error standard deviation: 1.6978335
Nominal Peaks Slope
The results for the interval prior to the fourth Kondratiev Wave trough support a rejection of the null hypothesis. The regression results are shown below.
Nominal Peaks Simple linear regression results:
Dependent Variable: IS Conflict
Independent Variable: Closest Distance (Years)
IS Conflict = 1.6751354–0.029363439 Closest Distance (Years)
Sample size: 122
R (correlation coefficient) = -0.15962545
R-sq = 0.025480284
Estimate of error standard deviation: 1.2927213
Nominal Peaks Slope
Interpretation of Results
As confirmed in the analysis, our primary regression signals no relationship between the two variables. However, the secondary regression produces a statistically significant indication that the slope of our regression is less than zero for the rate of intrastate violence initiation at the crest of Kondratiev Waves when compared to the other parts of Kondratiev Waves in years prior to the fourth Kondratiev Wave trough. This supports our hypothesis that Kondratiev Wave oscillations alter Perez’s technoeconomic subsystem in such a way that individuals’ cost-benefit analysis changes to favor rebellion more during Kondratiev Wave crests. We interpret these results to support the idea that violence, at least violence taking place solely within a state, has a cyclical component to it. Acknowledgement of a cyclical factor to intrastate violence is arguably a depressing claim to make. The implication is that, endogenously built in our nature as humans organizing in capitalistic society, we experience societal peaks in the propensity to commit violence against our neighbor.
However, there are two pieces of the analysis that complicate this picture. First, the R² value is incredible small, signifying that intrastate conflict initiation rates follow the trendline poorly. We interpret this value to mean that it is unlikely an intrastate conflict will start solely in the years of a Kondratiev Wave crest because of the oscillatory peak. Rather, the crest is merely one element that adds to heightened tensions. We must also accept that the results were not statistically significant for the dataset altogether. Clearly, something changes after the fourth Kondratiev Wave trough. This degrades the validity of our secondary regression because of the questions it force us to answer. We do not know why a shift occurs, but it is evident the Kondratiev Wave is not a completely static phenomenon throughout time. We have concluded this allows for two plausible scenarios. First, the correlation has ceased and will remain absent. Second, the Kondratiev Wave has undergone a phase shift and a new oscillatory pattern must be calibrated to accurately correlate with intrastate violence initiation rates. Understandably, this is a slippery slope. Regardless, we find this idea compelling when contextualized by scholars who have written about the modern profile of Kondratiev Waves. Though disputable, this possibility must be given some thought. Lastly, the absence of secondary variables in the regression does pose an empirical issue. Though we are unsure of the degree they could be removed, confounding variables are definitely present in this regression.
It is possible that we may learn a significant amount about long-run intrastate conflict trends by allowing for it to be partially a function of an endogenous cycle like a Kondratiev Wave. Further, what parallels can be drawn when we place the Kondratiev-Intrastate violence initiation interaction next to other cycles of conflict? For instance, states that have experienced a recent civil war have been empirically demonstrated to be at greater risk for future civil war. Collier eventually coined the term as a ‘conflict trap’ (2002). They note that civil wars can impede development or damage institutions critical to maintaining peace, resulting in an increased likelihood future conflict. A long-run perspective may see the alleviation of these issues, implying two possibilities. First, the general clustering of civil wars at the peak of Kondratiev Wave crests may partially be explained by the conflict trap. Alternatively, though we see progress across the 50-year interval in terms of development, it is possible that permanent damage has been sustained to the culture or its institutions that make civil war more likely in the foreseeable future.
The results of this study demonstrate one idea very clearly: there is an oscillation in the rate of intrastate violence initiation that occur in approximately 50-year intervals. However, it is less apparent if these cycles of violence are a direct result of Kondratiev Wave oscillations or due to an unknown, related phenomenon with the same interval frequency.
Two main concepts show promise as reasonable alternatives to our explanation of the results shown above. The first is that World Systems theory houses theoretical cycles that shift variables affecting intrastate violence initiation. These cycles terminate and begin at regular intervals with roughly the same wavelength as Kondratiev Waves. The second possible explanation is that a composite of smaller cycles may coincide to raise intrastate conflict initiations to the largest amplitude in 50-year intervals.
The most plausible and directly related alternative explanation can be found within the extensive literature available about World Systems theory and great power cycles. Raymond and Kegley offer Modelski’s theory of long cycles in world leadership as an explanation for fluctuations in rates of civil war initiations. Though Raymond and Kegley’s main contention is that civil wars become less internationalized as great powers enter the phase in which their power declines, we can extend this analysis to include the possibility of initiation rates fluctuating as well (1989). Linebarger’s concept of ‘rebel learning’ partially can explain when and how wars are initiated. He claims that wars are initiated partially as a function of inspiration and learning from other attempts to rebel (2016). Additionally, based on our reasoning in this paper, non-state actors may determine this is to be an opportune time to act because of great power states inadvertently exuding signals of decreased capacity to fight and control.
Besides than the explanation detailed above, few other cycles in the social science or STEM realms describe long-term cycles similar to the Kondratiev Wave. We believe it is possible that several shorter cycles may have their oscillatory peaks coincide every 50 years to produce an increased likelihood for intrastate conflict. The justification for this idea can be found in Korotayev and Tsirel’s spectral analysis of the Kondratiev Wave, which also illuminated existence of the shorter Juglar and Kuznets cycles as well. These two cycles were shown to boost the amplitude of Kondratiev Waves if all three oscillations coincided in the appropriate manner. Thus, it is possible the exact same phenomenon has occurred with several different cycles outside the realm of long-run macroeconomics. Climate cycles and electoral cycles show the most promise. Both have been individually linked to heightened instances of intrastate violence initiation. The concurrence of oscillatory peaks might produce results similar to what this paper has found.
Alteration to climate has become somewhat of a zeitgeist for the conflict research community as of late. These efforts are relevant and justified, but we must confine our analysis to purely repetitive variations of climate. Climate change is important and has been shown to increase the likelihood of intrastate conflict, but reflects a trend instead of a repeating undulation. However, components of climate that climate change research often is manifested through, such the El Nino pattern, can provide us with some utility. El Nino and its counterpart, La Nina, are variations in climate primarily located in the Tropic Pacific regions. These regular and noted instances of fluctuations in precipitation and temperature can be imagined as parallels to Kondratiev Wave peaks and troughs. The El Nino and La Nina pattern is admittedly placed in a smaller interval than Kondratiev Waves, but this is consistent with our alternative explanation of aggregated smaller patterns. Furthermore, a recent publication has also been produced that considers changes in climate to be adverse to stability and prosperity. The model the effects of climate change moving forward and claim that Africa is liable to see over 54% increase in the likelihood for intrastate conflict in the coming decades (Burke; Satyanath; Dykema; Lobell, 2009).
Electoral cycles prove to be another interesting candidate as a component to the ‘smaller and coinciding cycles’ theory. Gaubatz claims that there is evidence democracies initiate war more during the early stages of their respective electoral cycles. The later stages of electoral cycles are associated with less war (1991). While his analysis is centered on interstate warfare, it is possible that we may interpret this as an increased propensity for tolerance of general violence by the public. Framing our theory around the idea of lowered ratios of perceived costs to consequences around intrastate violence, it is possible that certain phases of the electoral cycle may signal the public’s increased tolerance to personally wage violence. This argument has some obvious flaws if we consider the average length of an electoral cycle and the fact that states’ electoral cycles are not usually synchronized. Furthermore, Gaubatz’s analysis is primarily centered on interstate war. Despite the weaknesses in this argument, Gaubatz’s electoral violence cycle provides an interesting framework to base alternative explanations from.
Lastly, there is the unlikely possibility that the results found in this paper were merely found by chance or are not indicative of a repetitive occurrence. However, rather than invalidating this study and its implications, the chance that these results are not due to fluctuations in Kondratiev Waves necessitates the duplication and elaboration of our study. This is a point that we will expand on in the section below dedicated to further research.
Limitations of Study
This study’s limitations also do not necessarily invalidate the findings, but instead illuminate effect these constraints will have on future significance and implications that the methodology and results have wrought.
Technical and data-oriented limitations were the most frequent issue with this study. The datasets used were not formatted with this type of analysis in mind, resulting in the hand-coding of each datapoint. While not a limitation in its own right, the process meant that extending the analysis to other potential subsets of variables increased the labor exponentially. Thus, the several confounding variables have not been removed and must be considered as well. The most critical variable not addressed is that of the core-periphery distinction. With Perez’s model claiming that crisis events are geared towards technoeconomic subsystem mismatches, we must acknowledge that socioeconomic structures are starkly different when examining core states and periphery states. It is possible our findings that Kondratiev Waves contribute to intrastate conflict applies more fittingly to one of the two. This issue is adequately addressed in the literature review, but the possibility that the distinction has affected our results is still present. Additionally, the sample is the population of states with instances of intrastate conflict. While this may be a strength in terms of statistical validity, this distinction limits our ability to conclude anything about future events for particular states. Each state is a unique case and has various conditions that affect intrastate conflict initiation. Our sample is not evenly distributed geographically or economically. The uneven occurrences of intrastate conflict in areas that are often less developed means that the sample is not equally applicable to a state chosen at random. With more instances of intrastate violence initiation being in the hundreds, we can observe that those instances are often unevenly distributed to certain regions or states. Thus, the conclusion reached in this paper may not apply identically to a state that has experienced two civil wars when compared to a state that has experienced 14 civil wars. This issue culminates into a realization that our results are perhaps fit to illustrate the relationship between Kondratiev Waves and intrastate violence initiation solely for states that experience a preponderance of rebellion to begin with.
Next, while the results were found to be statistically significant for the years prior to the fourth Kondratiev Wave trough, the R² value was prohibitively low. Thus, when combined with the irregular distribution of datapoints among states, the applicability of this study is called into question if the goal is to take an explanatory and anticipatory posture. Lastly, the results themselves present another limitation to the study’s applicability. The results no longer support the rejection the null hypothesis when including the most recent set of years. It is problematic to claim Kondratiev Waves are associated with an increase in intrastate conflict initiation if a shift has been shown to invalidate that assertion specifically for our most recent data.
Lastly, we admit the precarious nature of selecting a smaller timeframe to use as a secondary analysis. It is true that the results become less or more statistically significant when dividing up the dataset into individual Kondratiev Waves. However, we maintain a logical basis for the fourth Kondratiev Wave trough being the point of segmentation. While this satisfies the criticism to a certain degree, separate studies in which each of the Kondratiev Wave is analyzed would either supplement or challenge our claims in this paper well.
The relatively untouched and experimental nature of this study has profound implications for all future research that connects long-run macroeconomic cycles and intrastate conflict. We consider this study to be a rudimentary examination of an incredibly broad topic. Intrastate violence research has certainly become more popular in the last few decades, accompanied by a smaller sect of research that is specifically devoted to economic explanations for intrastate conflict occurrence. However, Kondratiev’s theory encompasses an interval that spans generations, making it both difficult to study and not necessarily the most appealing research topic because of its intangible character that can be attributed to the slow-burning oscillations. Therefore, our recommendation for future study regarding Kondratiev Waves and intrastate conflict is several-fold:
Further research must be done to determine whether the Kondratiev Wave continues to exist in its current form in the modern era. The investigation must also extend beyond merely identifying the existence of a modern Kondratiev Wave. Later scholars must explore causality to determine a more concrete mechanism for the model. Armed with an enhanced version of the model that applies a well-defined causal mechanism, the new model must be made to answer if the intervals are shortening and then explain why that occurs. After the previous goals have been accomplished, the model can then be extended into the future to anticipate upcoming long-run economic growth oscillations. The Kondratiev Wave may cease to function in such a way as to interact with Perez’s model of technoeconomic subsystems. If this is true, then the next logical area of study would be to examine whether the Kondratiev Wave affects intrastate violence in its altered position. While our reasoning for an interaction between intrastate violence and the Kondratiev Wave is grounded in socioeconomic organization, the Kondratiev Wave may continue to affect intrastate violence in ways that Perez’s model does not afford it.
Future researchers would be wise to examine what occurs at the troughs of Kondratiev Waves when growth is at its lowest level. While Perez’s crisis events mark a vital juncture for societal organization, it may be the case that Kondratiev Wave troughs affect intrastate violence in a different manner than examined in this paper. Perhaps the troughs lay create delayed structural issues that alter how societies utilize and distribute resources. Alternatively, the rivalry that existed under the US-USSR bipolar world order would certainly influenced intrastate war initiations as well. These states are also affected by Kondratiev Waves as well. It would be interesting to examine if fluctuations in growth that can be attributed to the Kondratiev Wave affect a state’s propensity to lend external support to non-state actors. The constitution and behavior of intrastate conflict must be continually examined to include new realities and developments. Intrastate warfare has been largely unchanged over centuries as a concept. However, we acknowledge the concept of a nation-state has not remained as stable and is also a fairly new concept. Scholars must uncover how intrastate warfare has and will change in the coming decades, along with the outstanding continuities. It is plausible the way that intrastate warfare is initiated and conducted will change on a fundamental level in the coming decades. Authors have recently began to publish about this possibility. Unprecedented levels of global connectivity, social change on an organizational level, and a new set of environmental constrains will certainly affect intrastate violence. The question is not so much if intrastate warfare will change. Rather, how will it change and will this affect the relationships described in this study that are crucial to allow for Kondratiev Waves to influence intrastate violence initiation?
We have demonstrated support for the argument that Kondratiev Waves are associated with an increased rate of intrastate conflict initiation. However, our research has left intrastate warfare duration and termination untouched. Perez’s model outlines a period in which there is a heightened level of social cohesion. In the context of intrastate violence initiation as an abstract function of opportunity costs, it would be worth examining if these periods illustrate a decreased propensity for intrastate violence because the social cohesion alters an individual’s attachment to the status quo and alleviates grievances that were more egregiously present during the crisis event.
A similar method of analysis to the one in this paper may be useful to repeat in a way that specifically analyzes core and periphery states separately. Given that institutional structures and baseline levels of economic growth are often vastly different between the two categories, it is possible these that the Kondratiev Wave and intrastate conflict interaction may be amplified or muted depending the state belonging to the core or periphery.
There certainly is no limit for the directions of further research on this topic. Due to the somewhat unexpected results in this paper, we argue the topic that is most deserving of additional research is how and why intrastate conflict and Kondratiev Waves may be transformed in the coming decades. Conflict anticipation is a vital motivations for research in this field. If these two phenomenon can be readjusted to reflect new realities, then the results found in this paper can be transposed to the new models. Hopefully, increased understanding will translate into actual muffled casualties and suffering in the real world and grow out of being solely theoretical arguments.
Contemporary democracies are not predisposed to think in terms of long-run consequences. This weakness is especially true of the United States, a global superpower that has been involved in combatting at least one insurgency at all times for the last 17 years. Leaders are often seduced to pursue what yields instant or near-instant gratification due to the struggles found in reelection and public approval. It is no wonder that Kondratiev’s theory was not studied more intensely the last century — a long-run theory means nothing to a senator who will be out of congress in 10 years and dead in 25.
The foundation for our research has been that long-run oscillations in certain factors affect individual and societal behavior. Perhaps partially explaining the deficiency in scholarship surrounding this subject, we must acknowledge that policy planning is focused on the most immediate stretch of time and neglects bracing for events that may transpire in the long-run. By acknowledging the hazards of insufficient foresight, strategic planning looks to be flawed. The Pentagon’s version of advanced planning is the Quadrennial Defense Review (QDR), a document outlining the US military’s priorities that is produced every four years. The conclusions reached in this study demonstrate the necessity for planning that extends further than a decade or two into the future, like each QDR does. While our research signifies that Kondratiev Wave crests do not affect the initiation of intrastate violence at a statistically significant rate after the 4th trough began in 1940, there are likely other long-run influences that have yet to be uncovered. Perhaps times that exhibit an increased likelihood for intrastate conflict initiation, leadership may designate a small and temporary increase in resource flows to CVE and stability programs.
We suggest this current lack of foresight is insufficient planning for a military, especially a superpower, like the United States, that is affected by conflicts abroad significantly more than other states. By aligning priorities in funding and strategic development to recognize threats far ahead of time, the military is afforded more flexibility and capacity to adequately meet these threats. Furthermore, patterns that become closer to stylized facts can be added into defense budget requests and deployments. The foresight will save lives and resources as the military acts to anticipate or mitigate violence. While the tone of the recommendations in this paper may appear somewhat realist and cynical, it is worth highlighting instances of civil wars the United States reacted too slowly to effectively stymie and the innumerable civilian lives that were lost as a result. We do not suggest that the Kondratiev Wave, in its current form or an altered one that is more accurately correlated with contemporary intrastate violence initiations, will be the sole factor that decides whether the US military correctly anticipates and responds to intrastate violence. However, the results show that intrastate violence is higher closer to Kondratiev Wave crests. An adjustment during these years may reduce the impact of such violence.
In examining the role Kondratiev Waves play in intrastate conflict initiation, we found moderate support for the claim that the crests of each wave are associated with greater intrastate conflict initiation than the rest of each wave. Contextualized by Perez’s technoeconomic subsystem paradigm and a rational choice model of war, we argue that intrastate conflict initiation increases during the Kondratiev Wave crests because the benefits that an individual acquires from continued participation in current the socioeconomic structure are at their lowest point relative to the costs from continued participation in the current socioeconomic. These crests correspond with a crisis event and promote (though they do not necessitate) conflict against the state and the institutions it upholds. Via a simple linear regression model, we illustrated this conclusion in an empirical manner. Although the results found in this paper became inconclusive after the 4th Kondratiev Wave trough, the favorable results found in the secondary regression are reason to believe the concept is something worth taking into consideration when attempting anticipate or explain intrastate conflict onset.
The existence of set 50-year intervals in which conflict oscillates, mandates that attention be given to long-run macroeconomic cycles in relation to intrastate conflict. The set interval does not appear to correlate after the 4th trough, but scholars have suggested a shift in the set interval may have occurred. The lack of validity in modern years does not imply that there is no correlation; it only implies there is no correlation with the current model. We have endeavored to show a link between Kondratiev’s writings and intrastate conflict, arguably succeeding. Researchers must account for this interaction whenever attempting to build a comprehensive model of intrastate violence. This necessitates further study of the interval and places a spotlight on the changing nature of intrastate warfare, long-run macroeconomic cycles, and their interaction.
It is unlikely there is any true static relationship between war, society, and economics. Kondratiev was tasked with making a value judgement on capitalism. He did not do this, but his findings have the possibility to contribute to peacekeeping efforts in future conflicts. We must further examine this prospect, preparing for the future to soften the harm societies can inflict upon themselves
Basu, Deepankar. (2017). Long Waves of Capitalist Development: An Empirical Investigation.
Boix, Carles. “Economic Roots of Civil Wars and Revolutions in the Contemporary World.” World Politics, vol. 60, no. 03, 2008, pp. 390–437., doi:10.1017/s0043887100009047.
Buhaug, Halvard, et al. “Square Pegs in Round Holes: Inequalities, Grievances, and Civil War.” International Studies Quarterly, vol. 58, no. 2, June 2013, pp. 418–431., doi:10.1111/isqu.12068.
Burke, M. B., Miguel, E., Satyanath, S., Dykema, J. A., & Lobell, D. B. (2009). Warming increases the risk of civil war in Africa. Proceedings of the National Academy of Sciences of the United States of America, 106(49), 20670–20674. DOI: 10.1073/pnas.0907998106
Caplan, Bryan. The Myth of the Rational Voter: Why Democracies Choose Bad Policies. Princeton University Press, 2008.
Caplan, Bryan. “Terrorism: The relevance of the rational choice model.” Public Choice, vol. 128, no. 1–2, 2006, pp. 91–107., doi:10.1007/s11127–006–9046–8.
Collier, P. and Hoeffler, A. (2002). Greed and Grievance in Civil War. World Bank, [online] CSAE WPS/2002–01. Available at: http://economics.ouls.ox.ac.uk/12055/1/2002-01text.pdf [Accessed 7 Jan. 2015].
Collier, Paul, and Nicholas Sambanis. “Understanding Civil War.” Journal of Conflict Resolution, vol. 46, no. 1, 1 Feb. 2002, pp. 3–12.
Connable, Ben, and Martin C. Libicki. How insurgencies end. RAND — National Defense Research Institute, 2010. Cornhill, C. H. “History of the People of Israel.” The Open Court, a Quarterly Magazine, vol. 11, no. 499, 12, 1897, pp. 733, American Periodicals.
De Greene, K B. “Towards New Conceptual Models of the Kondratiev Phenomenon.” Kondratieff waves, warfare and world security, edited by Tessaleno C Devezas, IOS Press, 2006, pp. 10–15.
De Groot, Bert, and Philip Hans Franses. “Common socio-Economic cycle periods.” Technological Forecasting and Social Change, vol. 79, no. 1, Jan. 2012, pp. 59–68., doi:10.1016/j.techfore.2011.06.006.
Dunaev, B. B. “Dynamics of Economic Cycles.” Cybernetics and Systems Analysis, vol. 53, no. 2, 2017, pp. 293–307., doi:10.1007/s10559–017–9929–3.
Elbadawi, Ibrahim, and Nicholas Sambanis. “How Much War Will We See? Explaining the Prevalence of Civil War.” The Journal of Conflict Resolution, vol. 46, no. 3, 2002, pp. 307–334. JSTOR, JSTOR, www.jstor.org/stable/3176229.
Fearon, James D. “Rationalist explanations for war.” International Organization, vol. 49, no. 03, 1995, p. 379., doi:10.1017/s0020818300033324.
Fearon, James D. and David Laitin 2003. “Ethnicity, Insurgency, and Civil War.” American Political Science Review, 97(1): 75–90.
Foster, John. “Joseph Schumpeter and Simon Kuznets: comparing their evolutionary economic approaches to business cycles and economic growth.” Journal of Evolutionary Economics, vol. 25, no. 1, 30 May 2014, pp. 163–172., doi:10.1007/s00191–014–0356–6.
Gallegati, Mauro, and Domenico Mignacca. “Jevons, Sunspot Theory and Economic Fluctuations.” History of Economic Ideas, vol. 2, no. 2, 1994, pp. 23–40.
Gaubatz, Kurt Taylor. “Election Cycles and War.” The Journal of Conflict Resolution, vol. 35, no. 2, 1991, pp. 212–244. JSTOR, JSTOR, www.jstor.org/stable/174145.
Goldstein, Joshua S. “Kondratieff Waves as War Cycles.” International Studies Quarterly, vol. 29, no. 4, 1985, pp. 411–444. JSTOR, JSTOR, www.jstor.org/stable/2600380.
Hoeffler, A. (2011), ‘Greed’ versus ‘Grievance’: A Useful Conceptual Distinction in the Study of Civil War?. Stud Ethn Nation, 11: 274–284. doi:10.1111/j.1399–6576.2011.01111.x
Jazayeri, Karen Bodnaruk. “Identity-Based political inequality and protest: The dynamic relationship between political power and protest in the Middle East and North Africa.” Conflict Management and Peace Science, vol. 33, no. 4, 2016, pp. 400–422., doi:10.1177/0738894215570426.
Kalyvas, Stathis N. and Laia Balcells 2010. International System and Technologies of Rebellion: How the End of the Cold War Shaped Internal Conflict The American Political Science Review Vol. 104, №3 (August 2010), pp. 415–429
Kegley, Charles W., and Gregory A. Raymond. “The Long Cycle of Global War and the Transformation of Alliance Norms.” Journal of Peace Research, vol. 26, no. 3, 1989, pp. 265–284.
Knoop, Todd. “Nineteen: Conclusions — What We Know and Do Not Know about Business Cycles.” Business Cycle Economics : Understanding Recessions and Depressions from Boom to Bust, ABC-CLIO, 2015, pp. 342–361. ProQuest.
Kondratieff, N. D., and W. F. Stolper. “The Long Waves in Economic Life.” The Review of Economics and Statistics, vol. 17, no. 6, 1935, pp. 105–115. JSTOR, JSTOR, www.jstor.org/stable/1928486.
Korotayev, Andrey V, and Sergey V Tsirel. “A Spectral Analysis of World GDP Dynamics: Kondratieff Waves, Kuznets Swings, Juglar and Kitchin Cycles in Global Economic Development, and the 2008–2009 Economic Crisis.” Structure and Dynamics, vol. 4, no. 1, 7 Jan. 2010.
Leontief, W. W. Domestic Production and Foreign Trade; the American Capital Position Re-Examined. 1953.
Linebarger, Christopher 2016. “Dangerous Lessons: Rebel Learning and Mobilization in the International System.” Journal of Peace Research 53(5): 633–647.
McCormick, Gordon H. 2003. “Terrorist Decision Making.” Annual Review of Political Science 6: 473–507.
Modis, Theodore. “A hard-Science approach to Kondratieffs economic cycle.” Technological Forecasting and Social Change, vol. 122, Sept. 2017, pp. 63–70., doi:10.1016/j.techfore.2017.04.015.
Mourao, Paulo Reis. “Does the Economic Cycle Change Our Prayers? A Discussion Based on Catholic Bulletins.” Kyklos, vol. 65, no. 4, 18 Oct. 2012, pp. 563–580., doi:10.1111/kykl.12006.
Niehans, J. “Juglars Credit Cycles.” History of Political Economy, vol. 24, no. 3, Jan. 1992, pp. 545–569., doi:10.1215/00182702–24–3–545.
O’Neill, Bard E. “The Nature of Insurgency.” Insurgency & terrorism, 2nd ed., Potomac Books Inc., 2006, pp. 15.
Perez, Carlota. “Structural change and assimilation of new technologies in the economic and social systems.” Futures, vol. 15, no. 5, Oct. 1983, pp. 357–375.
Perez, Carlota. Technological Revolutions and Financial Capital: the Dynamics of Bubbles and Golden Ages. Edward Elgar, 2014.
Narkus, Sarunas. “Kondratieff, N. and Schumpeter, Joseph A. Long — Waves Theory .” Kondratieff, N. and Schumpeter, Joseph A. Long — Waves Theory , University of Oslo, Forfatter, May 2012, Accessed 11 Feb. 2018.
Slantchev, Branislav L. 2003. “The Power to Hurt: Costly Conflict with Completely Informed States.” The American Political Science Review 97(1): 123–133.
Staniland, Paul 2014 Networks of Rebellion. Cornell University Press
Stewart, Frances. “Crisis Prevention: Tackling Horizontal Inequalities.” Oxford Development Studies, vol. 28, no. 3, 2000, pp. 245–262.
Thyne, Clayton 2006. “Cheap Signals, Costly Consequences.” Journal of Conflict Resolution 50(6): 937–961.
Tullock, Gordon. 2005. The Social Dilemma: Of Autocracy, Revolution, Coup d’Etat, and War. Ed. Charles K. Rowley . Vol. 8 of The Selected Works of Gordon Tullock. Indianapolis, IN: Liberty Fund.
Turchin, Peter. “Dynamics of political instability in the United States, 1780–2010.” Journal of Peace Research, vol. 49, no. 4, July 2012, pp. 577–591., doi:10.1177/0022343312442078.
Waldman, Thomas. War, Clausewitz and the Trinity. Ashgate Publishing Group, 2013.