Game Theory in Organized Crime

Game Theory in Organized Crime preview image

1 collaborator

Default-person Benedito Neto (Author)

Tags

game theory in organized crime 

Tagged by Qian li about 4 years ago

Visible to everyone | Changeable by everyone
Model was written in NetLogo 5.0.5 • Viewed 1353 times • Downloaded 52 times • Run 0 times
Download the 'Game Theory in Organized Crime' modelDownload this modelEmbed this model

Do you have questions or comments about this model? Ask them here! (You'll first need to log in.)


## WHAT IS IT?

This model simulates the relation between Government and Organised Crime from a micro-interactive point of view. The actions that the Government can take to fight against organised crime (or cooperate with it, and then share the benefits that arise from unethical agreements) depend on cultural aspects of the society studied, on the economic context in which global market is inserted, going through the State’s need to control this problem in a way that criminal organizations could not lead to society negative externalities (violence, constant clashes with government forces, parallel power’s competition), nor become more powerfull than the State itself (supposing a corruptive, or better saying, coerced Government that decides to cooperate with organised crime).

Recent studies in Korea conducted by professor Jonson Porteux verified that these kind of agreements with parallel power forces reduces the negative effects of illegal activies in the society – thus, organised crime is “better” than disorganized - it occurs because, by negotiating with these groups, Government can keep them under control and limit their influence over the welfare sphere. It also contributes to spur the lack of jurisdiction when a weak (politically or economically) government cannot itself take care of all society issues efficiently.

The purpose of this model is spur the studies of Professor Porteux in demonstrating graphically his studies. This is an environment coupounded by two criminal organizations and by government agents, who can work together with criminals or fight against them. It will be perceived that a non-cooperative behavior increases negative externalities over society and can amplify criminal organisations influence over welfare – and consequently reduce this.

## HOW IT WORKS

The initial interaction is non-cooperative. It will be perceived on “Criminal Activites” chart a continuous layout modification generated by upswings and downswings. Each downswing can be interpreted as a clash between a criminal firm and government, and each upswing is a criminal organization expansion over the market.

Generally, illegal firms will expand their power jointly in the market simulated, as a result of the government’s refuse in cooperating with them. When one of the firms is more influent than another, the weaker firm will cause a negative externality (faction clashes, increases in drugs traffic etc) to increase its power to become as powerful as the other – thus, with this simplification the social effects of competition between rivals to increase power can be well represented.

Always when government seems less propense to cooperate with criminals,illegal organizations will feel propense to expand their activities and influences.

In the “Wealth Flow” chart, the two mafious firms will compete head by head with government to economic power and political influence.

Turning the statepropensitytoCONTINUEtocooperatewithcriminals slider, a cooperative behavior starts and then changes in the graphic layouts will be perceived. Conflicts will be reduced, and also the competition for political influence. Government wealth will increase without harming criminal firms activities, just controlling their significance on well-being.

The statepropensitytoSTARTtocooperatewithcriminals slider must be calibrated before starting the interaction, and must not be turned during it, otherwise the results cannot be analysed uniformly in the temporal horizon.

## HOW TO USE IT

The interactive variables

Benefit-cooperation: This variable adjusts how much of profits are supposed the criminals would earn to from having cooperation with a corruptive government. It represents an increase in their usual earnings arising from cooperation, and varies from 0 to 1. It will be perceived that, in a non-cooperative interaction, that the higher the benefit of cooperation for criminals, the higher will be the negative externalities in society, thus, simulating the trade-off for the government in refusing a join in a cooperation.

StatepropensitytoSTARTtocooperatewithcriminals: This variable is a cultural variable, in which the effectiveness of the jurisdiction in relations among agents can be controlled. The stronger this variable is, the weaker a state jurisdiction will be since the beginning of the interactions.

StatepropensitytoCONTINUEtocooperatewithcriminals: This is a transitional variable, in the sense that you can control how your government would be willing to keep cooperating with organized crime indefinitely. This is independent if the initial condition at the beginning of the interactions. (Basically, it signifies that the government can get less or more opened to agreements with criminals in the course of the simulation).

Endowment?: This is an exogenous variable, which can be used or not to simulate the dynamics of the economy.

Endowment-rate: This variable controls the ‘flow of profits’ that a criminal can have. Basically, 0 simulates a market in relative good conditions and 10 a period of relative crisis.

Initial-number-criminals: Controls the number of criminals in the beginning of the simulation.

Initial-number-governmentagents: Controls the number of government agents in the beginning of the simulation.

Initial-state-strength: This variable controls simulated government’s wealth. A rich government can fight against organized crime more effectively than a less favored one. This can be tested by conducting the interactions.

Charts:

Criminal Activities: Controls the negative externalities the criminal firms can cause over society, like clashes or expansions.

Wealth Flow: Controls the money flow during the interactions among criminals and government agents, and also their increases or decreases in economic and political power.

In both charts, red and blue labels simulate the two criminal firms, and the black simulates the Government.

Monitors: The three monitors follow the aggregated wealth of the three agentes – both criminals firms and government. Always when the Government will be less oppulent than the criminals, a period of instability will be verified in the Criminal Activities chart, and its fading will be visualized in the Wealth Flow chart – it will signify that government will not be cooperating with organised crime, and then the firms will feel free to expand their illegal actions and cause negative effects in well-being, concomitantly getting a higher political influence.

## THINGS TO NOTICE

The model is quite sensible to changes, so must be accurately calibrated to generate the expected results.

It follows the CMM approach to simulate criminal behavior, in accordance to professor Porteux’s observations.

## THINGS TO TRY

• To have a proxy to reality, do not use more criminals than government agents as input. In a usual market, there are more government agents - as police, for example - than criminal people.

• This model has a more complex code than the macro interactive one, then its equilibrium requires more time on adjustments until the results searched can be obtained.

• Change the sliders during the interactions. From cooperation to non-cooperation and vice-versa. Quickly the changes in the chart layout will be observed.

• Depending on the calibration used, the ticks necessary to generate results and show the micro interactions and trends can vary. It’s strongly suggested to use also the macro interactions model.

## RELATED MODELS

"The Mafia Model - Interaction between police, mafia and storewoners" also named "Affecting Mafia With Social Norms"

http://ccl.northwestern.edu/netlogo/models/community/The%20Mafia%20Model%20-%20Interaction%20between%20police,%20mafia%20and%20storewoners

"Mafianomics"

http://web.econ.unito.it/terna/tesine/mafianomics.htm

## CREDITS AND REFERENCES

Adrian Haugen Ordemann - University Of Oslo

Benedito Faustinoni Neto - University of Sao Paolo

Based on the work of Professor Jonson Porteux - Hosei University

The CMM in chapter 10 in the book "The Economics of organized crime" by Gianluca Fiorentini and Sam Peltzman

Comments and Questions

Please start the discussion about this model! (You'll first need to log in.)

Click to Run Model

globals [endowment]

breed [criminals a-criminal]
breed [governmentagents a-governmentagent]
breed [criminals2 a-criminal2]
criminals-own [money]
criminals2-own [money]
governmentagents-own [money]
patches-own [countdown]

to setup
  clear-all
  ask patches [set pcolor green]
  if endowment? [ask patches
  [ set countdown random-normal 0 1   endowment-return                  
   set pcolor one-of [green red]                                          
  ]
  ]
set-default-shape criminals "criminals"                                        
create-criminals initial-number-criminals                                      
[
  set color black
  set size 2
                                           
  setxy random-xcor random-ycor                                           
  
]


set-default-shape criminals2 "criminals2"                                        
create-criminals2 initial-number-criminals2                                      
[  set color blue
  set size 2
                                       
  setxy random-xcor random-ycor                                           
] 

set-default-shape governmentagents "governmentagents"
create-governmentagents initial-number-governmentagents
[
  set color white
  set size 2
  set money initial-state-strength
 
  setxy random-xcor random-ycor
]


set endowment count patches with [pcolor = green]                                   
reset-ticks                                                                          
end 

to go
    if not any? turtles [ stop ]                                                 
  ask criminals [
    take-money               ; Should we not make a slider to adjust this threshold? So we can show that dependent on state-strengt you can get different equilibriums?               
    move                                        ; I think, according to the paper, that that can be of great value to look at different cases
    death 
    cheat
    multiply-criminals
    fight
    lie
    ] 
  ask criminals2 [
    take-money               ; Should we not make a slider to adjust this threshold? So we can show that dependent on state-strengt you can get different equilibriums?               
    move                                        ; I think, according to the paper, that that can be of great value to look at different cases
    death
    cheat
    multiply-criminals2
    fight2 
    lie
    ]
  
 ask governmentagents [
    move 
    cooperate 
    cooperate2
    catch-criminals 
    catch-criminals2                                    
    ]

  if endowment? [ ask patches [ endowment-return ] ]
  set endowment count patches with [pcolor = green]
 
 if benefit-cooperation < 0.5 [ask criminals [lie]]
 
 if benefit-cooperation < 0.5 [ask criminals2 [lie]]
 
  tick
end 

to take-money    
  if pcolor = green [
    set pcolor red
    set money money + 1 ]
end   

to move 
  rt random 50
  lt random 50
  fd 1
end 

to death 
  if money < 0 [die]
end 



;; to catch-storeowners
;;  let mafia-power sum ([money] of mafious in-radius 20) / 100  
;;  let ProbRefuse-myself sum ([money] of cops in-radius 20) / 100                     
;;  let prey one-of storeowners-here                   
;;  if prey != nobody                                                         
;;  [ask prey [ifelse (((ProbRefuse-myself  * ((storeowners-thrust-in-govs-ability-to-fight-mafia * police-power) / 2 ))) < mafia-power )
;;     [set money money - 3 ask patches in-radius 5 [endowment-return]] [set money money - 1 ask patches in-radius 2 [stop endowment-return]]]]
;;  ifelse (((ProbRefuse-myself * ((storeowners-thrust-in-govs-ability-to-fight-mafia * police-power) / 2 ))) < mafia-power )
;;   [set money money + 3] [set money money + 0]                   
;;  end                                                             


; to catch-criminals 
;  let market-profit sum [money] of criminals
;  let government-power sum [money] of governmentagents
;  let prey one-of criminals-here
;  if prey != nobody
;    [ask prey [ifelse market-profit > government-power and random-float 100 > 90   ;; According to paper
;      [set money money - (sum [money] of criminals-here)] [set money money + 0]]]     ; If true, criminals gets reduced money, if not there no change
;    ifelse market-profit > government-power [set money money + (sum [money] of criminals)] [set money money + 0]      
;   end                    ; If true the government gets the money of the criminals, if not, nothing happens

to cooperate 
  let market-profit sum [money] of criminals
  let government-power sum [money] of governmentagents
  let prey one-of criminals-here 
  if prey != nobody
    [ask prey [ifelse market-profit > government-power * ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) and ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) != 0
      [set money money + (sum [money] of criminals-here / 1000) * (1 + benefit-cooperation) ] [set money money + (sum [money] of criminals-here) / 1000 * benefit-cooperation / 2.5 ]]]   ; Funker ikke    ; If true, criminals gets reduced money, if not there no change
    ifelse market-profit > government-power * ( statepropensitytoCONTINUEtocooperatewithcriminals / ( 10 * statepropensitytoSTARTtocooperatewithcriminals) ) and ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) != 0
    [set money money + (sum [money] of criminals-here) * (1 + benefit-cooperation) * pi] [set money money + 0]         ; Funker
end           

to cooperate2 
  let market-profit2 sum [money] of criminals2
  let government-power sum [money] of governmentagents
  let prey one-of criminals2-here 
  if prey != nobody
    [ask prey [ifelse market-profit2 > government-power * ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) and ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) != 0
      [set money money + (sum [money] of criminals2-here / 1000) * (1 + benefit-cooperation) ] [set money money + (sum [money] of criminals2-here) / 1000 * benefit-cooperation / 2.5 ]]]   ; Funker ikke    ; If true, criminals gets reduced money, if not there no change
    ifelse market-profit2 > government-power * ( statepropensitytoCONTINUEtocooperatewithcriminals / ( 10 * statepropensitytoSTARTtocooperatewithcriminals) ) and ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) != 0
    [set money money + (sum [money] of criminals2-here) * (1 + benefit-cooperation) * pi] [set money money + 0]         ; Funker
end           

to catch-criminals 
  let prey one-of criminals-here
  if prey != nobody
    [ask prey [ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) = 0 and benefit-cooperation = 0 and random-float 100 > 90   ;; According to paper
      [set money money - (1 + benefit-cooperation) * (sum [money] of criminals-here)] [set money money + 0]]]
    ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) = 0 and benefit-cooperation = 0 and random-float 100 > 90 [set money money + (sum [money] of criminals-here)] [set money money + 0]
end 

to catch-criminals2 
  let prey one-of criminals2-here
  if prey != nobody
    [ask prey [ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) = 0 and random-float 100 > 90   ;; According to paper
      [set money money - (1 + benefit-cooperation) * (sum [money] of criminals2-here)] [set money money + 0]]]
    ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) = 0 and random-float 100 > 90 [set money money + (sum [money] of criminals2-here)] [set money money + 0]
end 

to cheat
  let prey one-of governmentagents-here
  if prey != nobody
    [ask prey [ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) = 0 and random-float 100 > 90 ;
     [set money money - (sum [money] of governmentagents-here)] [set money money + 0]]]                   ;always when they start to become more powerful
    ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) = 0 
    [set money money + (sum [money] of governmentagents-here)] [set money money + 0]
end 

to endowment-return                 
  if pcolor = red [                 
    ifelse countdown <= 0                        
    [set pcolor green
      set countdown endowment-rate ]           
   [set countdown countdown - 1 ]             
  ]
end 

to multiply-criminals
    
    let my-money sum [money] of criminals-here
    let enemy-money sum [money] of criminals2
    if my-money > 100 * (enemy-money / (benefit-cooperation + 0.01))    and  ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) <= (benefit-cooperation * 10) or count criminals < count criminals2  [hatch 1 rt random-float 360 fd 1 set money money - (sum [money] of criminals-here)] 
end                               

to multiply-criminals2
    
    let my-money sum [money] of criminals2-here
    let enemy-money sum [money] of criminals
    if my-money > 100 * (enemy-money / (benefit-cooperation + 0.01))    and ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) <= (benefit-cooperation * 10)   or count criminals2 < count criminals [hatch 1 rt random-float 360 fd 1 set money money - (sum [money] of criminals2-here)] 
end                               

to fight
  let my-profit sum [money] of criminals
  let enemy-profit sum [money] of criminals2
  let prey one-of criminals2-here
  if prey != nobody
    [ask prey [ifelse my-profit < enemy-profit and random-float 100 > 90   ;; According to paper
      [set money money - (sum [money] of criminals2-here)] [set money money + 0]]]
    ifelse my-profit < enemy-profit [set money money + (sum [money] of criminals2-here) ] [set money money + 0]
end 

to fight2
  let my-profit sum [money] of criminals2
  let enemy-profit sum [money] of criminals
  let prey one-of criminals-here
  if prey != nobody
    [ask prey [ifelse my-profit < enemy-profit and random-float 100 > 90   ;; According to paper
      [set money money - (sum [money] of criminals-here)] [set money money + 0]]]
    ifelse my-profit  < enemy-profit [set money money + (sum [money] of criminals-here) ] [set money money + 0]
end 

to lie
 
  let government-power sum [money] of governmentagents
  let enemy-profit sum [money] of criminals
  let enemy2-profit sum [money] of criminals2
  let prey one-of governmentagents-here
  if prey != nobody
    [ask prey [ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) != 0 and benefit-cooperation < 0.5 and (enemy-profit + enemy2-profit) * benefit-cooperation = initial-state-strength * government-power and random-float 100 > 90 ;
     [set money money - (sum [money] of governmentagents-here)] [set money money + 0]]]                   ;always when they start to become more powerful
    ifelse ( statepropensitytoCONTINUEtocooperatewithcriminals / 10 * statepropensitytoSTARTtocooperatewithcriminals ) != 0 and (enemy-profit + enemy2-profit) * benefit-cooperation = initial-state-strength * government-power and random-float 100 > 90
    [set money money + (sum [money] of governmentagents-here)] [set money money + 0]
end 

There is only one version of this model, created about 4 years ago by Benedito Neto.

Attached files

File Type Description Last updated
Game Theory in Organized Crime.png preview Preview for 'Game Theory in Organized Crime' about 4 years ago, by Benedito Neto Download

This model does not have any ancestors.

This model does not have any descendants.