Power and Leadership: A Complex Systems Science ApproachPart I—Representation and Dynamics

Yaneer Bar-YamNew England Complex Systems Institute, 277 Broadway, Cambridge MA 02139

(Dated: October 21, 2018)

Historical social narratives are dominated by the actions of powerful individuals as well as compe-titions for power including warfare, revolutions, and political change. Advancing our understandingof the origins, types and competitive strength of different kinds of power may yield a scaffoldingfor understanding historical processes and mechanisms for winning or avoiding conflicts. MichaelMann introduced a framework distinguishing four types of power: political, military, economic,and ideological. We show this framework can be justified based upon motivations of individualsto transfer decision making authority to leaders: Desire to be a member of a collective, avoidingharm due to threat, gaining benefit due to payment, acquiring a value system. Constructing modelsof societies based upon these types of power enables us to distinguish between social systems anddescribe their dynamics. Dynamical processes include (a) competition between power systems, (b)competition between powerful individuals within a power system of a society, and (c) the dynamicsof values within a powerful individual. A historical trend in kinds of power systems is the progres-sive separation of types of power to distinct groups of individuals. In ancient empires all formsof power were concentrated in a single individual, e.g. Caesar during the Pax Romana period. Inan idealized modern democratic state, the four types of power are concentrated in distinct sets ofindividuals. The progressive separation of the types of power suggests that in some contexts thisconfers a “fitness” advantage in an evolutionary process similar to the selection of biological organ-isms. However, individual countries may not separate power completely. The influence of wealth inpolitics and regulatory capture is a signature of the dominance of economic leaders, e.g. the US.Important roles of political leaders in economics and corruption are a signature of the dominance ofpolitical leaders, e.g. China. Ideological leaders dominate in theocracies, e.g. Iran. Military leadersdominate in dictatorships or countries where military leaders play a role in the selection of leaders,e.g. Egypt.

I. INTRODUCTION TO POWER AND GROUPACTIVITY

Social power is the ability to control or influence thebehaviors of others. History is often described throughthe interaction between individuals exerting power overone another. Power relations among a set of people deter-mine whether they are considered as a group or politicalentity. Sociologists have been concerned with the under-lying nature of power, how it arises, how it affects thosethat have it and those that don’t, its roles in regulat-ing relationships between people in groups ranging fromhouseholds to global civilization, and categories of power[1–7].

Michael Mann [8–11] developed a framework in whichthere are four types of power: ideological, economic, mili-tary and political. In his approach these are loosely con-nected societal networks of control. Ideological powerresults from the ability to control meaning in society be-cause people have a need to define meaning, norms, andritual practice. Examples include religions as well as sec-ular ideologies such as Marxism. Economic power derivesfrom control over the satisfaction of subsistence needsthrough the extraction, transformation, distribution andconsumption of the objects of nature. Political powerderives from the usefulness of centralized, institutional-ized, territorialized regulation of many aspects of socialbehavior. Military power derives from the necessity oforganized physical defense and its usefulness for aggres-

sion, and more specifically concentrated lethal violence.Mann provides an extensive historical discussion of ex-amples of how these forms of power occur and relate toeach other over history.

Here we formalize the framework developed by Mannby showing that it is a “universal” framework for rep-resenting power due to the inherent structure of indi-vidual motivation. Mann did not find social theory dis-cussions of motivations starting from basic human drives(e.g. sexual fulfillment, affection, health, physical exer-cise and creativity, intellectual creativity and meaning,wealth, prestige, power, etc) useful. Our approach moredirectly considers the mechanisms and dynamics of powerin group activities. The objective is to understand howmultiple individuals together perform activities in whichan individual, or small group of individuals, determineswhat is done. Motivation at the individual level is di-rectly understood from the perspective of the individ-ual who yields rather than the individual who exercisespower. The individual who yields power must have amotivation for the cognitive act of yielding control overactions, expecting a better condition from doing so thanwould otherwise occur. We understand these forms ofpower as motivated respectively by: the perceived bene-fit from membership in a group—political control; theavoidance of direct harm—military power; the receiptof direct benefit—economic power; and change in val-ues self-consistently motivating that change—ideologicalpower. On the one hand, this discussion relates to the

arX

iv:1

811.

0289

6v1

[ph

ysic

s.so

c-ph

] 7

Nov

201

8

2

traditional political science of elites that are able to influ-ence or determine the behavior of groups. On the otherhand, this is an extension of a complex systems character-ization of the mechanisms by which collective behaviorsarise [12, 13].

II. FROM INDIVIDUAL TO COLLECTIVEBEHAVIOR

Social systems are built out of coordination of individ-uals into collective action. One of the important mecha-nisms for the creation of such behaviors is through con-trol by one individual over the actions of others. This is atraditional mechanism by which collective action arises,distinct from self-organized behaviors such as those ofswarms and flocks which can arise without an individualexercising control over the collective. The latter has beenthe subject of much of complex systems study of collec-tive action due to its limited treatment in prior work.Here we extend complex systems concepts to the moretraditional topic of control and power.

Examples of behaviors that involve multiple individu-als performing the same tasks in unison include marchingsoldiers in a Roman legion, or coordinated actions in afactory. A central question is identifying what model ofan individual is needed in order to identify the mecha-nisms of collective action. Our objective is not to identifythe actions that are performed but rather the mecha-nisms of control over actions.

Considering an individual in isolation may not serveas a basis of description of collective behaviors in manycontexts. In this case, however, there is a meaningfulprimitive concept: an individual in isolation performsactions for self-benefit. Among these actions are foodgathering or shelter construction. How such individuallydirected actions are transformed in a social context tobecome collective behaviors, that may or may not benefitthe individual, other specific individuals, or society as awhole, is the question we are addressing.

Mechanisms of control and the actions that they giverise to may or may not be evaluated as positive or neg-ative within a value system (i.e. power may be used toconstructively coordinate or to coerce and exploit). Thisis not the central question we are discussing, though someidentification of the way types of power are interpretedin value systems is appropriate once we have identifiedthe mechanisms.

In considering the structure of power systems it isuseful to identify different types of power. Distinguish-ing a few types cannot be a complete specification oftheir diversity, but can provide a basis for discussion ofthe dynamics of power and the evolution of social sys-tems including, but not limited to, governance struc-tures. Our analysis will show that the four types of poweridentified by Mann [8]—political, military, economic andideological—are immediately justified in this context.

The development of a theoretical framework enables

inferences about various properties of power and its dy-namics. Given different forms of power, their interactionsover time are an important part of the dynamics of powerhistorically. The consequences of dominance of one indi-vidual or the dominance of one type of power over otherscan be seen historically.

An example of such a dynamic is embodied in the ex-pression ‘power corrupts,’ which interestingly suggeststhat power can be used for either collective or individ-ual benefits but generally will result in the individualones predominating over time. In considering that thereare multiple forms of power, we can reasonably hypothe-size and observe historically that when one type of powerdominates other types over time more negative conse-quences for the public arise as well as outsized benefitsfor those in power. By comparison, a balance over differ-ent kinds of power, and dynamic power shifting, yieldsbenefits for the public. We can understand the reasonsfor this observation from an analysis of dominance anddynamics.

This paper is organized as follows: In Section III weframe social power as the transfer of authority over de-cision making. Section IV provides an abstract formalintroduction to the four types of power. Section V de-scribes political power, Section VI military and economicpower, and Section VII ideological power. Section VIIIdiscusses the role of institutions in power. Section IX andSection X describe the representation of power graphi-cally. Section XI introduces a notation for power systems.Section XII describes the use of historical data to discusspower systems. Section XIII describes a selection of his-torical examples. Section XIV describes the dominanceof power types. Section XV describes the dynamics ofpower. Section XVI provides concluding summary.

III. FROM PHYSICAL TO SOCIAL POWER:FORCE ACTING OVER DISTANCE TO

TRANSFER AUTHORITY

It is interesting to begin by considering the distinctionbetween power as it is understood in physics and in so-cial systems. Power in physics is a force exercised overa distance. We can understand the power to change themovement of an object in this fashion. The greater thepower, the larger the change in momentum (a change inspeed or direction of motion of an object) that can beachieved, where other forces are not present. Physicalpower thus reflects an ability to determine the behav-ior of the object. There are circumstances where suchpower is also exercised over a person, i.e. direct forcechanges the motion of an individual. For example, onecan exercise power over another by pushing them in acertain direction. While this mechanism of power doesoccur, it is significantly different than the way power insocial systems is often described. In social systems, it istypically assumed that each individual continues to beautonomously directed, but the control over an action

3

that is performed can be transferred from the individ-ual who is performing that action to another individualwho wields control. This concept of the transfer of au-thority over decision making is central to the theoreticalframework of power we describe here.

When we consider the role of power in social collec-tives, power reflects a transfer of authority over decisionsabout actions by many individuals to one or a few indi-viduals, or alternatively to a social abstraction such aslaws or institutions. Such a transfer of authority givesrise to large scale social actions. The scale of those ac-tions is to be considered the subject of a dynamic the-ory of the exercise of power by various authorities. Ourframework provides a basis for such a dynamic theory.

IV. FORMAL FRAMEWORK FOR POWER

In presenting a scientific analysis of a social constructsuch as power there is a need to provide formal justifica-tion for the model constructed as well as a presentationthat can be understood at an intuitive level. These arelinked to observations that begin from quite general onesand then proceed to progressively more detailed valida-tion. This section provides a more abstract formal ap-proach to arrive at the same framework for power as thesubsequent sections and can be omitted at a first readingby those wishing for the more intuitive approach.

We consider an individual as having the ability to makedecisions about multiple possible actions, i.e. to chooseover time a sequence of actions to perform of many possi-ble such sequences. We consider the choices by the indi-vidual to be guided by “motivations” which are distinctfrom individual decision making about specific actions intheir generality, i.e. motivations guide many decisionsand their modification changes the overall pattern of de-cision making. The distinction between motivation andspecific decision making is only in the abstraction of themodel of decision making, i.e. a motivation is an aspectof the causal reason for many different decisions ratherthan a single one.

We define power as the transfer of decision making overone individual’s actions from that individual to anotherindividual (a leader). We are not primarily concernedabout the mechanism of that transfer of power, i.e. theforms of communication that are present to facilitate orenable it. We will be particularly concerned with thecase where the leader has power over multiple individualsrather than a single individual.

In building our framework of power, we consider howmotivations of an individual lead to power over the ac-tions of that individual by another. Such a transfer ofpower can itself be considered an action and thereforeshould be understood from motivations of the individ-ual. Human motivations are often limited in scientificdiscussions to striving for self-benefit (i.e. economic ra-tionalism). We consider a more general concept of anobjective oriented motivation—actions that are causally

connected to valued conditions, i.e. values—with self-benefit as an important and common example of suchvalues due to its self-consistency through selection forpersistence through relevance for survival. We allow forself-benefit or measures of achievement of values associ-ated to an action to be measured both in a positive andnegative sense, with terminology such as harm or cost forthe negative direction.

Beginning from a model of an individual who performsactions for self-benefit, decision making is motivated bythe individual’s understanding of benefits and costs tothe individual. This decision making can be influencedif the transfer of power to the leader has the ability tochange the apparent payoffs of the set of actions that arepossible for an individual to make.

We therefore consider four ways a transfer of powercan be motivated. First, by the existence of an underly-ing expectation of improvement of benefit from the verynature of a transfer of power. Second and third, by theleader directly increasing the apparent cost or the appar-ent benefit to the individual of particular action options,leading to those actions being avoided or chosen due tothose changes in payoffs. Fourth, by changing the eval-uations of the payoffs by the individual, e.g. throughchanging values. These four options map onto the fourdifferent forms of power that we will consider (political,military, economic, and ideological). We note that themotivation for the fourth (power to change values) canarise self-consistently from within the values themselves,and has a potential to be inherently recursive, a key prop-erty of this mechanism for motivation.

V. POLITICAL POWER AND GROUPCOLLECTIVE ACTIVITY

In order to understand power we must also understandwhy group collective activities may serve purposes in so-cial systems. In some contexts, there exist beneficial ac-tivities that can be performed by joint action of a group,which individual action cannot perform. Examples in-clude successful hunting of large animals (i.e. attacksby multiple individuals may kill a large animal undercircumstances in which an attack by one individual can-not), time bounded searching (i.e. search and rescue),and dominance in military conflict. More generally, itis widely believed that a large variety of problems canbe solved by putting an individual “in charge” of theactions of multiple individuals. This belief is consistentwith complex systems theory. When one individual isable to determine the behavior of multiple individualscomprising a collective, this enables the behavior of thatcollective to have a scale and complexity that may beneeded for the successful performance of a task [12, 13].We note that it is possible to show that the benefitsof such individual authority are bounded under highlycomplex conditions where effective distributed control ismore beneficial. However, there are also circumstances

4

under which the expectation that putting an individualin charge will have benefits is justified. Under these cir-cumstances it makes sense for an individual to desire tobe part of a group that transfers authority over individ-ual decision making to one individual, i.e. an individualchooses to be a member of a group rather than “going italone.” For our purposes it does not matter if this resultsfrom evolutionarily derived (instinctive) behavior or a so-cially learned one (conscious or not). We call this typeof transfer of authority for an implicit purpose of collec-tive (and individual) benefits “political power.” Thus wesee that it can be important to have a leader to performeffective collective action.

While collective action may have inherent benefits, in-dividual differences may give rise to different levels of de-cision effectiveness for political leaders. Thus, it may beimportant which leader is given that power. How a groupdetermines which individual is given political power is akey aspect of how we understand group activities gener-ally and systems of collective governance specifically.

We associate our understanding of political power withthe conventional concept of political leaders. This is con-sistent with the observation that the mechanism of selec-tion of such leaders can vary and does not have a specificorigin that is characteristic of other forms of leadership.Other justifications for this association will arise fromspecific examples.

VI. MILITARY AND ECONOMIC POWER

Two other kinds of power can be readily identified asthe result of threat or reward, the proverbial stick andcarrot incentives. In the former, a direct threat results inan individual choosing to transfer authority to avoid po-tential harm. In the latter, transfer of authority is doneto achieve a direct benefit of a specific action. Theseare distinct from the general transfer of authority repre-sented by political power in that they arise from specificindividual assessment of sufficiently well defined negativeor positive consequences of alternative options.

An institutionalized version of power that arises fromthreats involves a military or police force, so we termit “military power.” Actions that arise from associatedbenefits rather than threats are often described as volun-tary and contrasted with threats as being coercive, andare considered as the subject of economics. We thereforeuse the term “economic power.” Our focus is not on alleconomic transactions, but rather on transactions thatinvolve actions by one individual (labor) in response tobenefits (wages).

We note that the decision of an individual to accepta benefit depends on the relative benefit associated withalternatives that are available. Considering the relativebenefit makes the distinction between voluntary and co-ercive concepts less clear than is often assumed in eco-nomics. I.e. if the only alternative to a transaction isnegative the decision to transact is coercive in much the

same way as a direct threat. For example, the alterna-tive to wages may be starvation and therefore the choicemade is essentially based upon a threat. Considering anindividual who has control over a critical resource, i.e.water, such control can enable economic power that iscoercive. Thus, the distinction between voluntary andcoercive arises because of the nature of the alternatives.When there are, in some sense, adequate alternativesthe exercise of economic power is different from coer-cion. Military power is by definition dependent on anadverse alternative; economic power is dependent on ad-verse alternatives only in so far as there are not betteralternatives. In the study of economics it is assumed thealternative conditions that might be coercive are not theresponsibility of the benefit provider. Accordingly, it isnot the provider that is the coercive agent, which distin-guishes economics from coercion. At the same time, theanalysis of behavioral choices and actions of an individ-ual must consider the set of options that are available tothem and the causes of those options.

VII. VALUES AND IDEOLOGICAL POWER

It is apparent that underlying the exercise of economicpower are the values of the individuals to whom ben-efits are offered. Values are embodied in the “utility”of something to an individual as an essential primitiveconcept in economic theory. Values are also essential topolitical power, as the acceptance of leadership, and morespecifically the choice of a leader to whom authority isgiven, depends on the values of those who are voluntarilytransferring authority to the individual who becomes theleader. Perhaps less obviously, values also underly theexercise of military power. While the threat of pain ordeath is an important motivation for accepting a transferof authority, individuals may refuse to heed threats, maywithstand pain, or even accept death, based upon values.The ability of different individuals to adopt different val-ues and change them over time, and crucially to transferauthority over the setting of those values, leads to an ad-ditional form of power, “ideological power.” Ideologicalpower in its purest form does not act directly to controlactions but rather controls values that in turn controlthe decisions about actions. Ideological power interactsin important ways with other forms of power. For ex-ample, by changing values it can reinforce or undermineeconomic, military or political leaders. By this definition,ideological power can be seen to include both ideologicalvalue systems and the individuals who exercise authorityover/within those systems. It is also apparent from thedefinition that ideological power encompasses religiousvalue systems but also systems that are not generally in-cluded as religious, including patriotism and secular hu-manism. Value system leaders include all of those whoprovide narratives of meaning and significance about theworld. In a secular context, this includes authors, re-porters, entertainers, producers and teachers. There are

5

Economic

+

–Military

Political

Ideological

FIG. 1: We identify four types of power that enable one in-dividual to control the actions of others. Economic and mil-itary power arise from rewards and threats, political powerfrom the expectation of group benefit from central decisionmaking, and ideological power from transfer of authority overvalues.

embedded values in any fiction or non-fiction story andreporting, including what is important and how it is de-scribed.

VIII. INSTITUTIONS AND CONVENTIONSWITH POWER

Laws, institutions and other social conventions shouldbe understood to serve two (equivalent) roles in ourframework. First, they serve as the “system of power”within which power is exercised. Second, they can beunderstood to circumscribe the power individuals exer-cise at a particular time. In effect their presence assignspower to historical figures and processes which are ab-stractions of individual authority. The mechanism of au-thority of those abstractions must be embodied throughindividuals with current power whose authority can beunderstood in relation to those abstractions (e.g. spe-cific individuals who have interpretive or enforcing roles)as well as the public. The values of the public in as-signing authority (i.e. the authority they confer on ab-stractions and current leaders) constrains the degree towhich leaders can exercise power in relation to those ab-stractions. In an important sense, laws, institutions andsocial conventions as values are incorporated in the reli-gion/ideology of the populace.

Thus, for example, religious (and other) leaders maybe constrained in their actions by codified religious val-ues. Political, military and economic authorities (as wellas religious ones) may be constrained by legal systems,institutions and conventions based upon historical prece-dent. When leaders test the constraints of those cod-ified systems, potentially violating or changing them,the response of other leaders or the populace determineswhether they are able to do so.

IX. GRAPHICAL REPRESENTATION OF ANINDIVIDUAL

Fig. 1 provides a schematic diagram of an individ-ual showing political, military, economic and ideologicalpower as consistent with a simple model of individual be-havior consisting of (1) a tendency to transfer authorityto others (political power), (2) a desire for benefits (eco-nomic power), (3) avoidance of harm (military power),and (4) a system of values, over which authority can alsobe transferred (ideological power). This elementary di-agram of how the decision making of an individual iscontrolled and the actions that result can be used to il-lustrate various types of power systems. A power systemcharacterizes the way aggregation of actions occur wheredifferent groups of individuals are controlled by compet-ing powerful individuals, potentially with different typesof power.

X. GRAPHICAL REPRESENTATIONS OFAGGREGATION (MULTISCALE ANALYSIS)

We can use the individual level representation toconstruct collective representations of social systems inwhich power is exercised. An example is shown in Fig.2 in which a military leader, e.g. a dictator, uses mil-itary power to exercise control over the populace. Theschematic not only shows that power is exercised overthe populace but explicitly includes the role of individ-uals in the military to control the public, and the con-trol of the leader over the military that gives rise to thecontrol of the leader over the public. As is shown inthe schematic, the relationship between the leader andthe military may involve multiple channels of power, in-cluding values (ideological, i.e. loyalty, obedience), po-litical (authority transferred for group benefit, i.e. forthe military group or for the country as a group), mili-tary (i.e. coercive, threats that might also originate fromother members of the military [not shown separately]),or economic (i.e. wages). Note also that economic powerover the leader is exercised by the populace as a whole,i.e. through the flow of taxes to the leader. This power isnot the same as that of an individual exercising power dueto its distribution through the populace (but suggests theimportance of such reciprocal power, especially if wealthbecomes concentrated, or in the event of collective ac-tion). The complexity of the diagram illustrating a realworld system that involves military power is consistentwith a need for embodying the power in relationshipsbetween individuals used to create that power. Thus,saying a military leader exercises power is not enough,the military force through which the power is being ex-ercised is part of the representation of that power. Avariety of other kinds of power structures can be simi-larly illustrated by such diagrams (e.g. Fig. 3).

6

Public

PublicMilitary/Police

Leader

FIG. 2: Schematic of a social system in which military powerdominates. The leader uses its control over a military or policeforce to exercise power over the populace, extracting resourcesfrom the populace that it uses to reward and thus control themilitary force.

PublicMilitary/Police

Political LeaderMilitary Leader

Ideological Leader

Economic Leader

FIG. 3: Schematic of a social system in which leaders witheach of the four powers play a role. Many but not all of theroles that power can play are shown. Military forces are shownin addition to the military leader. However, in a more com-plete figure there would also be various political operatives,corporate managers and clergy and lay leaders.

XI. NOTATION FOR REPRESENTATION OFSOCIAL POWER SYSTEMS

In this framework, a society is described by a set ofpower relationships that comprise a kind of network. Wecan define a society as such a network that does notdecompose into independent subgroups. A quantitativerepresentation of a society is a list of individuals or groupswith a specified number of members and the associatedindividuals or groups that exercise power over them:

(n1, name1,name1,p,name1,m,name1,e,name1,r)(n2, name2,name2,p,name2,m,name2,e,name2,r)...(nk, namek,namek,p,namek,m,namek,e,namek,r)

(1)

where namei is an individual’s name or a group name

with number of members ni. Each of the other names ina list is the identifier of an individual or group exercisingpower over that individual, i.e. a leader of a particulartype, in the order of political (p), military (m), economic(e), or ideological (r). A null or self-reference of a leaderrepresents the case where that type of leader is not exer-cising power over the individual, or does not exist. Theproperties of the social system depend on the numbersof members of groups and the structure of the networkof power links. For analysis of the structure the labelingby names is not essential and may be replaced by ordinalnumbers that index that individual or group.

The formalism can be extended to include laws andother forms of constraints on leaders, i.e. constitutions,social conventions, religious documents, or other codifiedguides. The simplest way to include them is to namethem and to allow the number of individuals associatedwith that element of the system to be zero. This is sim-ilar to the inclusion of extrinsic forces on the system ingeneral models of influence networks [14, 15]

For example, a single leader with a military based dic-tatorship with 1,000 military members and simultane-ously religious authority over a million citizens might berepresented as:

(1, 1, 1, 2, 3, 1)(1000, 2, 1, 2, 1, 1)(1000000, 3, 1, 2, 1, 1)

(2)

Based upon historical information about the role of Ro-man Caesars in political, military, and economic leader-ship, this might be considered a simplified model of thePax Romana system (see below).

A dictatorship with a separate religious leadership with100 clergy might be represented as:

(1, 1, 1, 3, 4, 2)(100, 2, 1, 3, 1, 2)(1000, 3, 1, 3, 1, 2)(1000000, 4, 1, 3, 1, 2)

(3)

The notation can be extended to represent systems inwhich more than one individual or group exercises par-tial power over a particular individual or group, by mak-ing the substitution namek,s → {namek,s,i, wk,s,i} with sin the set {p, e,m, r}, i an index, and wi the weight ofthe respective influencer with partial power. This wouldthen be a network representation of how individuals areinfluenced by multiple other individuals or groups. Thenetwork representation we are identifying has additionalstructure compared to a standard influence network [14–16] by virtue of distinguishing the different categoriesof influence, and allowing nodes to represent aggregatedgroups rather than individuals. The representation is dif-ferent from that generally considered in network theoryof generalized mutual influence with a specified structureof the influence network. It has more distinct structuralelements, but collapses the number of individuals of thesociety and the relationships among them onto a con-

7

cise and simpler form. This simplification, in which indi-vidual agents have power over groups, may be useful inanalysis.

The framework developed thus far in this paper de-scribes a system of roles rather than of individuals withinthose roles. Missing is the mechanism by which indi-viduals are placed into those roles. Those mechanismsare often integral to the power system itself. Indeed, intraditional political thinking the definition of a politi-cal structure is often based upon that mechanism. Thusa dictatorship or a representative democracy might inprinciple (though not usually in practice) differ only inthe way the individual in power gains that power ratherthan the powers that that individual exercises. We cangeneralize the framework we have discussed to include acharacterization of the processes by which individuals at-tain power. There is, however, enough to discuss in thecontext of the power structure itself for a first article sowe defer much of the discussion of the processes by whichpower is attained to a follow-on article, only mentioningit when essential. Instead we focus on the dynamics ofpower systems, the way that the structure of power insociety, as described by the representation we are using,changes over time.

The dynamics of power systems can be considered interms of the representation by identifying (1) illustrativeexamples of power systems and their historical trendsand (2) dynamics of power systems through (a) compe-tition between systems, (b) changes of the structure ofa system by changes in the types of roles present (i.e.whether a single individual plays multiple power roles),or the power relations among powerful individuals, and(c) changes in the values of individual that plays a cer-tain role. Constructing examples requires the ability torefer to history, a process fraught with scientific issuesbut one which cannot be avoided in this context.

XII. USE OF HISTORICAL CONSENSUSNARRATIVES AS SCIENTIFIC DATA

Collecting data about social systems for scientific anal-ysis presents many challenges. Among these is the imple-mentation of controls on the objectivity of observationaldata. In order to discuss our models of power it is help-ful to have information to compare with across historicalperiods. Data sources are often limited to observationsby individuals that are then transmitted through a chainof communications to the present. Establishing historicaldata “facts” for scientific analysis is therefore problem-atic. However, where information affects many individu-als, the historical communication of information is morereliable than the communication of information that per-tains to individual actions that are less influential. Thisreliability arises because of the redundancy of the infor-mation, which promotes valid transmission through un-reliable communication channels [17]. In this case, thecommunication channel is the historical transmission of

social information. We therefore consider consensus his-torical narratives that describe events or individuals as totheir influence on many people as having an appropriate,if limited, level of validity for inclusion in scientific anal-ysis. It should be apparent that the validity of even suchevents may be challenged. The opportunity for such chal-lenge exists and notwithstanding the possibility that in-dividual events may be found to be incorrectly reported,a scientific theory may be compared to existing histori-cal narratives. We will proceed to consider the way thathistorical power systems can be interpreted within theproposed power framework. Mann’s four volume treatise[8–11] can be consulted for a much more extensive dis-cussion of the interplay of the four types of power aroundthe world and through history.

XIII. THE EXERCISE OF POWER BYLEADERS

A power system, by definition, means that leaders havethe ability to control what other individuals in the sys-tem do. The actions of individuals that are controlledcan serve many different purposes. Actions that can becontrolled can be considered a “resource” like land orraw materials. Leaders will control the actions of othersto achieve what they themselves consider benefits. Thusthe value system of the leaders is an essential attributeof leadership and its role in society. The premise of indi-vidual motivations implies a central question is whetherthe values of the leader result in self-serving benefits orbenefits to those who are controlled by them, or othergroups. The way leaders are selected plays an importantrole in the leader values that are likely to be present.Achieving benefits that are desired by a leader, includ-ing self-serving benefits, is a motivation for individualsto become leaders. Given multiple leaders, the ability tocontrol is limited by or shared with other leaders. Thislimits the ability of one leader to achieve what they be-lieve is of benefit. Therefore, one of the likely goals of aleader is to increase their power, so as to increase theirability to serve what they consider to be a benefit. Asdiscussed later, this is a generic part of the dynamics ofpower systems whether or not the leader is self-serving.

XIV. ILLUSTRATIVE HISTORICALEXAMPLES OF POWER SYSTEMS

The four different types of power can interact with eachother in different ways in a specific system. Whethertypes of power are combined together within an indi-vidual or distributed among distinct individuals and theway powerful individuals exercise power over each other,so that one type of power might dominate others, distin-guishes different systems.

The existence of different forms of power must be com-bined with social separation of these powers in order for

8

them to be recognizably distinct at an aggregated level.Historically, they are not, however, completely separatedbut in different systems they become segregated in vari-ous ways. Understanding the structural forms of powerdistribution and their dynamics is important scientifi-cally and relevant historically. Perhaps a good way tothink about the way this works is like identifying pri-mary colors in a picture. The elementary nature of theseforms of power allows for concentrated patches but doesnot preclude various forms of mixing that are consideredby Mann as networks of power.

Here we limit our discussion to a few observations andgeneral statements of relevance to historical dynamics ofpower systems. One observation that appears supportedby the historical record is that power has become disag-gregated over time in large countries compared to ancientRome and other early empires. A common origin of thegrowth of ancient empires was through military conquest.This naturally led to military leaders that dominate otherforms of power. By contrast, in modern military dictator-ships, ideological power (often religious) and to varyingdegrees economic power are often separated as they werenot in ancient ones.

In the ancient Roman empire during the Pax Romanaperiod, a Caesar exercised all forms of power, i.e. po-litical, military, economic and ideological leadership. Inparticular, Caesar was considered a god during this pe-riod. At the beginning of this period, the classical Romancivilization went from a republic dominated by militaryaction to a society with essentially three groups [8]: theemperor, soldiers, and the mass of the people—includingslaves (an estimated 20-40% of the population at the be-ginning of this period) and nominally free peasants thatmerged together into serfdom. A small number of people,the leadership of the military, might also be considered adistinct group. The budget (combining that of the stateand that of the emperor, which were not substantiallydistinct) was 70% for military and 15% for the dole tothe Roman public, presumably the minimum that couldbe given and maintain order. In contrast, US militaryspending is 20% of total federal spending / 12% of fed-eral, state and local spending. The Pax Romana periodwas as pure a version of a coercive society as may bepossible with human beings. Its success in space-timeextent can be understood from the stability of such asystem given the value system of the time.

The advent of the major religions led to partial,though surely not complete, separation of ideologicalpower from other forms of power in many areas of theworld. Later secular ideologies, including communism,secular humanism, and more generally patriotism, whenthey arose became more closely associated with politicalpower. Theocracies embody the possibility that ideolog-ical power becomes dominant, exercising also all otherforms of power to a greater or lesser degree. This is ap-parent in the periods of Christian Inquisition and theCrusades, the Islamic conquest and caliphates, and themodern state of Iran as key examples. The dominance of

religious leaders over other forms of power can be distinctfrom military leaders exercising religious authority dueto the origins and primary mechanisms of their power.Communism suppressed competing ideologies includingreligions. Modern governance structures that appear tobe either absent a religion or having adopted pluralis-tic separation of church and state may be better charac-terized as variations on patriotic secular humanism. Inthis context, the separation of church and state enablessufficiently compatible religious systems to coexist withoverlapping roles in determining values.

Confucianism provides an important example for ourpurposes. Mann’s characterization[8] suggests that“Confucianism’s role is thus largely one of morale boost-ing: it introduces no principles of ideological transcen-dence.” Our analysis suggests a different perspective fo-cused on the role of the value system itself. Confucianismhas a transparent and self-declared purpose of shapingthe values of leaders away from self-interest and towardbenefit to others, as well as loyalty of subjects. Accordingto its principles, having virtuous rulers is central to stablegovernance. In this it directly targets the essential role ofleadership values: When leaders adopt values that servethe public, they are supported by the public. Leaderswould adopt it because of the public loyalty it inherentlyas well as explicitly promotes. When the public consid-ers leaders as not fulfilling those values, leaders wouldbe replaced. Indeed, leadership changes did occur andConfucianism dominated Chinese values for 2,000 yearsuntil the advent of Communism. Confucianism also em-bodied an extreme version of a codified value system,where written abstract ideas are the essential version ofthe ideological power. The common knowledge of theseideas and their transmission through educational insti-tutions is, to first order, the extent of institutionalizedpower structure.

Since the industrial revolution and through the 20thcentury, the relationship between economic and politicalpower was the subject of much of the global ideologicalconflict, with different degrees of dominance of politicalleadership over economic resources and thus economicpower as reflected in capitalism, socialism, and commu-nism. These ideologies are in part descriptions of powersystems, though the relationship between the ideologicalvalues and the reality of the power system effects may notbe direct or complete because of limited understandingof the real world consequences of their implementation.That people value and expect certain outcomes of powersystems does not make them true properties of those sys-tems.

In an idealized well functioning modern democracy,the four domains of power are considered to be sepa-rated. Each of the powers dominates in its own sectorof public activity, political leadership making policy, mil-itary and police leadership ensuring security, economicleadership making production and investment decisions,and religious leadership promoting altruistic and patri-otic values. Political power exercises limited control over

9

economic power through regulation, and over militarypower through control over resources, and also throughultimate command and the declaration of war.

While separation of the domains of power is consid-ered the ideal, the exercise of power of economic leadersover political power through regulatory capture is widelyconsidered to be found in capitalist systems that em-phasize independence of economic activity from politicalcontrol, including in the US. Corruption is widely con-sidered to be present in the places where political powerhas greater control. We note that US policy has longopposed corruption. This may seem to be paradoxical ascorruption is superficially similar to the kind of influencethat economic interests have over the US government.However, our analysis shows that there is a difference.Influence buying or regulatory capture occurs when eco-nomic power is being exercised over political power. Incontrast, corruption reflects dominance of political powerover economic power. Hence, economic leaders do not likecorruption. The difference is subtle but real. The prac-tical reason they are different is that regulatory captureenables “special interests” to exercise control over pol-icy, while corruption enables a political power to extractmoney for a specific favor that need not provide systemicbenefits, i.e. it is part of business activity (a kind oftax) rather than part of business policy. When politicalpower is dominant (China today), political leaders are theprimary beneficiaries of economic activity (i.e. wealth).When economic power dominates, economic leaders arethe primary economic beneficiaries and they have effec-tive control over economic policy (US today).

When one power dominates, that power gains author-ity over other forms of power, determining public decisionmaking (political power), coercively promoting their con-trol (military power), extracting wealth from and control-ling economic activity (economic power), and promotingvalues that favor themselves (religious power). Undersuch conditions the boundaries between different kindsof power become blurred.

Moreover, when one type of power becomes strongerthan other types of power, individuals whose leadershiprole originates from one source of power become leaderswith other types of power. Thus, when economic powerdominates, wealthy individuals become political leaders.When political power dominates, political leaders becomeleaders of economic institutions, i.e. corporations. Thesetwo directions are apparent in the examples of the USand China today [18, 19]. Consequences are similar formilitary and ideological/religious leaders when they aredominant.

It is important to emphasize that the populace exer-cises power in every system and this power is manifest indiagrams (Figs. 3 and 4) that show the populace exert-ing power over leadership. Collectively, the power of thepopulace is strong, generally stronger than any leader-ship, but individuals or even groups of the populace haverespectively less power if they do not act in a unified way.

XV. DOMINANCE OF TYPES OF POWER

More systematically, we can list the effects of domi-nance of one kind of power over another:

4 Where military power dominates over

5 political power: we have dictatorship, policestate, loss of rule of law

5 economic power: we have coercive exploita-tion, extortion (e.g. organized crime)

5 religious power: we have generals as icons orgods (e.g. Rome)

4 Where economic power dominates over

5 political power: we have regulatory capture,exploitation of natural resources and exploita-tion of workers (e.g. low wages)

5 military power: we have coercive employmentincluding prison labor and slavery

5 religious power: we have aspiration for wealthand/or acceptance of subservience, work isdone for hoped for (prospect of) money andnot for money

4 Where political power dominates over

5 military power: we have wars of conquest

5 economic power: we have corruption

5 religious power: we have patriotism, suppres-sion of other religions

4 Where religious power dominates over

5 military power: we have persecution, inquisi-tions and religious wars

5 economic power: we have cult behavior

5 political power: we have religious law, prohi-bition, witch hunts (e.g. McCarthyism)

Modern examples of relative but not complete domi-nance of each of the four forms of power can be found inEgypt for military power, US for economic power, Chinafor political power, and Iran for religious power.

XVI. DYNAMICS OF POWER

There are three kinds of dynamics of power: competi-tion between power systems and their specific instantia-tions, competition of leaders for power including aggrega-tion and inter-leadership relationships, and an individualleader’s transformation of values that affects how theyallocate the resources they have control over.

While there are distinct aspects of these three formsof dynamics, there are also ways in which they all reflectthe competition between leaders. Leaders compete for

10

power within a power system, as well as to gain controlover others that are currently part of a different powerstructure. The latter can be reflected in conflict betweentwo systems with the same type of power, or two systemswith different types of power systems. Individual leaderschange their values and associated behaviors in ways thatdepend on how vulnerable or secure their control is inrelation to other leaders. Still, dynamics may not beentirely reflective of leader competition for power, as, forexample, power systems can compete even when they arenot subjects of the same individual leaders.

A. Dynamics of power systems

One of the central aspects of the historical narrativeis the changes that occur in power systems as reflectedin the prevalence of system types. While our analy-sis of power is not exactly the same as the traditionalcategories of political systems, changes in political sys-tems are changes in power systems. Thus the histori-cal changes of prevalence of monarchies, dictatorships,democracy and communism reflect a competition be-tween power systems. This competition occurs throughgradual change within a country, conquest by one coun-try of another, as well as by revolution and mimicry ofpower systems by one country of others.

The outcomes of competition between power systemsreflect the advantages that one power system has over an-other. Intuitively a historical progression among powersystems arises because they provide increasing competi-tive advantages, a kind of evolutionary dynamic. Com-petitive advantage may arise from two different sources,which are not independent: advantage to the populace,hence their selection by the populace, and advantages incompetition at the collective scale, in one or more of thecategories of military, economic or religious conflict, asreflected in the mechanisms by which such conflicts arewon.

In direct conflict between power systems where leadersof the same type and their followers compete, the winneris determined by qualities that can be readily identified,though other aspects of the system may play a signif-icant role. For example, conflict between militaries isgenerally considered to be won by the scale of militaryforces as measured by manpower and firepower. How-ever, factors such as military technology and productioncapacity may play a key role and unconventional war-fare points to other factors of importance in particularcontexts. Conflict between religions is won by prosely-tization and stubbornness (as well as ethnic conflict /massacres and conquest of territories). Conflict betweeneconomic systems is won by resources and their effectiveallocation (as well as innovation). Conflict between polit-ical power systems is won by popularity contest (as wellas political maneuvering).

When we consider the relative advantages of differenttypes of power systems, we should recognize that evolu-

tionary advantages are context and tradeoff dependentand therefore involve subtle considerations and carefulanalysis. This is manifest in biological evolution throughthe diversity of biological organisms that are present inthe world today. For example, it is commonly stated thata faster running cheetah has evolutionary advantage overa slower one, hence the high speed they achieve. How-ever, not all animals are as or even at all rapidly running,nor are some able to run at all, and some can fly fasterthan a cheetah can run. This points to the context spe-cific nature of selective forces. We may infer that suchcontext specificity is also present in the evolution of socialsystems and their power structures. As a first step, how-ever, we can identify some advantages of power systemsin relation to each other by straightforward considera-tions. These advantages are to be understood as subjectto clarification of the conditions under which they apply.For example:

. Military dictatorships provide order and are thuspreferred over disorder, i.e. anarchy, as an absenceof a power system. Dictatorships will be selectedfor over anarchy under conditions where large scaleresources are available, including those obtained byconquest, and other external forces are not disrup-tive of the power system. This is similar to multi-cellular organisms, which have advantages over sin-gle celled organisms in the presence of large scaleresources (e.g. food) that can be consumed, oneof which may be mutual consumption, and becauseof the homeostatic environments that are formedwithin them.

. Monarchies are advantageous over dictatorshipsdue to reduced conflict over leadership transitions.This will be selected for under conditions of multi-generational persistence of the power system.

. Democracies and Communism as political systemsare advantageous over monarchies due to the ad-vantages associated with selection of leadership val-ues and capabilities. They will be selected for un-der conditions in which leadership decisions mustrespond to an elevated level of internal complex-ity of the society (note however the limitations ofhierarchical representative systems [13]).

. Free market based Democracies are advantageousover systems that exercise price and productioncontrol (e.g. Dictatorships and Communist sys-tems) due to the advantages of economic coordi-nation for complex economic activities.

. In general, we can argue that competition amongcountries due to system capabilities and effective-ness and for popular support tends to aggregatepower where large scale coordinated actions areneeded, and disaggregate power through specializa-tion of leadership and increased complexity of ac-tions in the face of increasing complexity in actions.

11

The latter is similar to the evolutionary develop-ment of cognitive capacities in animals throughneural systems as being an evolutionary advantagein the context of environmental complexity of deci-sion making.

Competition between groups involved in direct con-frontation may reflect competition between power sys-tems, when the power systems are different. It mayalso reflect competition between different instances of thesame power system. In the latter case it may in part re-flect competition between leaders for control over largerpopulation groups, a topic that is further discussed in thenext section.

Finally, we note that power systems may change notbecause of direct competition but rather because ofchanges in the conditions. Thus, discovery of naturalresources and technological changes can change the re-sources that are present in the system. The changes inresources then affect power structures. For example, theindustrial revolution in changing the means and organi-zations associated with economic wealth also created newpower structures.

B. Dynamics of leadership

Evolutionary dynamics are only one aspect of the dy-namics of power systems, just as evolutionary dynamicsare only one aspect of the dynamics of biological organ-isms. For clarity, among the others in biology are devel-opment, aging, damage, recovery, metabolic, cellular andmolecular processes.

Key to the dynamics of power systems are internalcountry competition and cooperation between leaders.Leadership competition tends toward consolidation ofpower as leaders seek increasing control as a means toachieve their objectives. Note that this internal dynamicof power aggregation is generally counter to the evolu-tionary dynamic that seeks to distribute that control foreffectiveness of the system. Aggregation of power is lim-ited by death, and thus mechanisms of inheritance are akey aspect of how power aggregates over time.

In the process of power consolidation by leaders, dif-ferent types of leaders can employ different methods. Wecan identify these methods from their type of power. Ineach case there is a mechanism of direct power use to takecontrol, and a mechanism of cooperation that can resultin power sharing, mutual strengthening and dependency.Specifically:

. Political leaders gain control through popular sup-port for subverting or conditionally legitimizingother leaders.

. Military leaders gain control by threatening or byproviding security for other leaders.

. Economic leaders gain control by buying out or pro-viding resources for other leaders.

. Religious leaders gain control by proselytization ofother leaders or by advocating on their behalf toothers.

When a particular type of leader gains control they notonly change the balance of power between leaders, theychange the nature of the power system by strengtheningthat type of power. Note that in all cases the differencesbetween different types of methods used may be drivenby values as well as circumstances.

Thus, for example, where leaders cooperate, economicpower supports military power through taxation as partof social values promoted by religious leaders reflected inlaws established by political power. Military power sup-ports political power through preventing anti-leadershipor revolutionary activity, as well as defending against for-eign occupation and other threats. Ideological leaderspromote values of leaders promoting economic, securityand other benefits for the public, as well as values of thepublic (and leaders) that reinforce support of leaders. Po-litical leaders implement laws and allocate resources forthe public, as well as for economic growth and militarydefense, reflecting values promoted by religious leaders.

There is an important distinction between leaders thatcooperate with each other for mutual support and leadersthat cooperate in support of the system of power itself.Military leaders might promote particular economic orpolitical leaders, but this would undermine the structureof an economic system in which economic competitionis the mechanism of promotion of economic leaders, anda political system in which political leaders are selectedby public opinion and thus motivated and having capa-bilities in the promotion of public good. Similarly, eco-nomic leaders might promote individual political leadersby financial support, but this undermines the structureof the power system in a context in which political lead-ers are selected because of motivation and benefits thatthey provide to the public.

As another example, where military power competeswith economic power, military power can threaten eco-nomic power to obtain resources, or economic power canbuy out military power, i.e. gain authority over it byproviding support. Either way money flows to militarypower, and coercive control to economic power. If theextent of such competitive interaction increases, militaryforces may gain extensive control over wealth, or eco-nomic forces may gain extensive control over military ac-tions to control workers by coercion instead of payingfor labor. In this context we can see how values (ideol-ogy/religion) are an important aspect of the competitionfor power. For example, military values may limit theextent to which military leaders seek wealth. On theother hand, if it favors wealth, then military power maythreaten economic power to obtain dictatorship controlover economic resources. The competition also occurs inthe context of popular support. For example, militarypower and economic power compete if they are both try-ing to gain political power.

We can see that the different kinds of power cooper-

12

ate or compete in different structures. In a cooperativecontext:

. Military power provides for security and enforce-ment of laws, i.e. social order, which promotes theframework in which economic, political and reli-gious power is exercised,

◦ it enforces rules of ownership and protects thewealthy from crime,

◦ it provides security for the political leadersagainst opponents,

◦ it provides security for religious leaders,

◦ it provides security for the public.

. Political power provides

◦ laws that protect military, economic and reli-gious leaders,

◦ laws that enable economic leaders to makeeconomic decisions and exploit resources,

◦ subsidies for economic activity, for the mili-tary, and for the populace.

. Religious power provides values that promote

◦ public respect for economic, political and reli-gious leaders and their decisions,

◦ public actions in support of collective causes,

◦ distribution of resources for the well being ofthe society and the public in particular.

. Economic power provides management of resourcesand labor and coordination of processes

◦ for the public good,

◦ in support of military actions and religiouscommunications,

◦ in support of political figures reflecting suc-cessful outcomes of policy decisions that pro-mote economic growth and social benefits.

In each case we can replace positive with negative interac-tions for systems with antagonistic interactions. Antago-nistic interactions undermine leadership. Undermining isdistinct from, but may be part of, interactions that leadto takeover by one power of another.

C. Values of leaders and its dynamics

Power can serve for the benefit of the group, includ-ing the individuals that are subject to power. It is alsoapparent that power can serve for the benefit of the in-dividuals who are wielding power. Which is the case?

Decisions by an individual in power may lead to vary-ing degrees of benefit for the collective and for the indi-vidual in power. Values of a leader determine whether

their decisions are intended for the benefit of themselvesor for the collective that empowers them (or anothergroup). Thus the benefits of power can either serve anindividual or the collective. While intentions and out-comes are not always aligned, under conditions in whichconsequences of actions are clear, intent corresponds tooutcomes: Capability and values combine to determinesocietal and individual benefits.

Quite generally, the values of an individual affect (1)what he does for himself and (2) what he does for others(others may be distinguished in group categories). Foran individual who wields power and controls actions ofothers and receives resources from others the values ofthat individual affect (1) what he uses/takes (allocates)for himself and (2) what he uses/gives (allocates) to oth-ers (in group categories). This includes both immediateuse and allocation for future use.

Thus, all types of leaders may and often do use powerfor self-benefits, including self-benefits that are consid-ered to be counter to the idealized values of the powersystem. Self-benefits may be considered through biolog-ical and social needs and desires. This is manifest infinancial gain and associated luxuries, and sexual miscon-duct as reflected in “scandals.” Note that this includesreligious leaders even when the values they profess andespouse for others contradict such actions.

We can identify a theory of individual transformationof values in the expression “power corrupts.” This state-ment embodies a dynamic concept. Specifically, thatthere is a change in individual values as a result of thewielding of power. The change shifts the behavior ofthe individual toward serving him or herself rather thanothers. Implicit in this dynamic is the existence of al-ternatives, i.e. that an individual wielding power maynot be serving him or herself. We see that this simpleexpression captures a remarkably rich set of concepts. Itsuggests that persistence of power leads progressively tomore self-allocation and use of power for self-benefit.

XVII. CONCLUSIONS

We have identified four types of power based upon theunderlying structure of human motivation and charac-terized the historical nature of power systems and thedynamics of power in terms of them. Societal benefitsfrom establishing leadership for collective action can beidentified. However, the essential paradox of transfer ofpower to a leader is that the purpose of the transfer ofpower as far as a group is concerned is for benefit to themembers of the group, but the opportunity is present forthe leader to exercise power for self-benefit. Power dy-namical processes include (1) the dynamics of change ofthe relative benefit assignment to leader and group, (2)the competition and cooperation among different lead-ers, including different types of leaders, and (3) compe-tition among power systems. It can be expected thatthe dynamics of individual leader values favors increas-

13

ing self-benefit over time (power corrupts), the dynamicsof power shifting may under some conditions favor groupbenefit but also power aggregation, and the evolution-ary dynamics of groups and power systems at least undersome conditions favors distributing power and group ben-efits, though group benefit may not equally serve every

group member. Historical processes appear to result in ashift toward distributed power systems, though individ-ual countries today continue to be dominated by specifictypes of power including political (e.g. China), economic(e.g. US), military (e.g. Egypt) and religious (e.g. Iran)leaders.

[1] T. Hobbes, Leviathan or The Matter, Forme and Powerof a Common Wealth Ecclesiasticall and Civil (1651).

[2] R. P. J. French Jr, B. H. Raven, Studies in Social Power ,D. Cartwright, ed. (Institute for Social Research, AnnArbor, MI, 1959), pp. 150–167.

[3] J. K. Galbraith, The Anatomy of Power (Houghton Mif-flin, Boston, 1983).

[4] M. Mann, The Autonomous Power of the State: Its ori-gins, mechanisms and results, European journal of soci-ology 25, 185 (1984).

[5] J. R. French, B. Raven, The Bases of Social Power, Mod-ern classics of leadership 2, 309 (2001).

[6] D. Keltner, D. H. Gruenfeld, C. Anderson, Power, ap-proach, and inhibition., Psychological review 110, 265(2003).

[7] S. T. Fiske, J. Berdahl, Social power, Social psychology:Handbook of basic principles (2nd ed. pp. 678–692 (2007).

[8] M. Mann, The Sources of Social Power, Vol. 1: A historyof power from the beginning to AD 1760 (1986).

[9] M. Mann, The Sources of Social Power, Vol. 2: The riseof classes and nation states, 1760-1914. (1993).

[10] M. Mann, The Sources of Social Power, Vol. 3: Globalempires and revolution, 1890-1945 (2012).

[11] M. Mann, The Sources of Social Power, Vol. 4: Global-izations, 1945-2011 (2012).

[12] Y. BarYam, Dynamics of Complex Systems (Westview

Press, Reading, MA, 1997).[13] Y. BarYam, Multiscale variety in complex systems,

Complexity 9, 37 (2004).http://onlinelibrary.wiley.com/doi/10.1002/cplx.

20014/abstract.[14] D. D. Chinellato, I. R. Epstein, D. Braha, Y. Bar-Yam,

M. A. de Aguiar, Dynamical response of networks underexternal perturbations: exact results, Journal of Statis-tical Physics 159, 221 (2015).

[15] D. Harmon, M. Lagi, M. A. de Aguiar, D. D. Chinellato,D. Braha, I. R. Epstein, Y. Bar-Yam, Anticipating eco-nomic market crises using measures of collective panic,PloS one 10, e0131871 (2015).

[16] R. Axelrod, The dissemination of culture: A model withlocal convergence and global polarization, Journal of con-flict resolution 41, 203 (1997).

[17] C. E. Shannon, A mathematical theory of communica-tion, Bell System Technical Journal 27, 379423 (1948).

[18] Wikipedia, Anticorruption campaign under Xi Jinpinghttps://en.wikipedia.org/wiki/Anti-corruption_

campaign_under_Xi_Jinping.[19] Wikipedia, Regulatory Capture

https://en.wikipedia.org/wiki/Regulatory_

capture.