Refugees, Resource Scarcity, and the Ethnic Balance of Power in Lebanon

The literature has posited three primary mechanisms

• Refugees introduce arms and ideologies conducive to violence (Salehyan and Gleditsch 2006)
• Refugees can exacerbate competition over exceedingly scarce resources (Bohnet 2012)
• Refugees can change the ethnoreligioous balance of power between rival groups (Forgsberg 2013)

Refugees add to demographic pressures that result in increased competition over

• Access to clean water, food, and shelter/housing
• Employment opportunities
• Access to government services

Refugees can also contribute to changes in the 'ethnoreligious balance of power' by

• Contributing 'demographic power' to one group over its rival(s)
• This obfuscates the demographic equilibrium between rival groups
• Which results in uncertainty over the distribution of power between these groups
• Under these circumstances increased conflict between between formely rivalrous but peaceful groups emerges

• The majority of approaches involve analysis at the country level
• No substate analysis outside of Africa has been undertaken
• Testing of the ethnoreligious balance of power mechanism has only been undertaken either at the country level of analysis or through small-n comparative case studies

• GIS methods of exploratory data visualization
• Qualitative Comparative Analysis (QCA)
• Spatial Regression (Count and Probit Models)

• The Syrian Civil War has emerged as one of the bloodiest domestic conflicts in recent decades
• Refugees have been fleeing in record numbers to all of Syria's neighbors (with the exception of Israel)
• While Jordan has managed to maintain stable domestic conditions, Iraq, Lebanon, and Turkey are increasingly spiraling towards conflict
• The focus of this project is Lebanon

• Most dire refugee crisis of all neighbors
• Shares a similar sectarian tapestry with Syria
• In Iraq, where ISIS has managed to wrest control of over one fourth of the country, conflict has become so pervasive and deadly that the relationship between the geography of refugees and violence in that country is highly endogenous
• Lebanon is the only state to block the establishment of formal refugee camps by the United Nations

Qualitative Comparative Analysis (QCA) will be performed on Lebanon's 26 administrative districts

• Unit of Analysis: Lebanese administrative 2 districts (Caza)
• DV: Number of conflict events per district
• IV: Number of Syrian refugees, sectarian tensions, income level, population density

• Data on number conflict events (DV) has been collected for each district from March 2013 to May 2015
• Conflict events refer to any events that involve the use of force or violence related to armed conflict, which includes arrests of suspected militants, bombings, intercommunal violence, state repression, shootings, assults, etc.

• First introduced by Charles Ragin (1987, 2000)
• Applies rules of logical inference associated with qualitative designs to data that will be examined with quantitative techniques of deductive reasoning
• Permits the study of outcomes that result from set-theoretic expectations
• Formalizes Mill’s “method of difference” by applying either Boolean algebra or fuzzy logic to sets with the aim of determining which combinations of conditions result in a study’s outcome of interest

• Equifinality
• Appropriate for research designs with a limited number of cases
• Relies on case-specific knowledge
• Compliment to mainstream statistical approaches, not an alternative to them

• Lacks standard techniques of inference (none with small-n)
• Lacks an appreciation for temporal or sequential dynamics
• High degree of sensitivity to the coding choices of the researcher
• High degree of sensitivity to measurement error
• Checks of robustness help to overcome some of these issues

• Sets <-> Variables
• Outcome Set <-> Dependent Variable
• Conditions <-> Independent Variables
• Cases <-> Observations

• Identify outcome of interest and the cases that exemplify outcome -positive cases
• Identify all 'negative cases'
• Propose a number of hypothesized causal conditions relevant to this designated outcome
• Calibrate set memberships
• Construct “truth table”
• Apply boolean alegbra or fuzzy set theory to identify necessary and/or sufficient conditions for outcome
• Assess causal paths with parameters of fit

• Set membership must be calibrated according to one of three principles
  • crisp set membership
  • fuzzy set membership
  • multivalent set membership

• Sorts cases by combinations of causal conditions they exhibit
• All combinations of conditions are listed (even those without empirical instances)
• Assess the consistency of the cases in each row with respect to the outcome
• In crisp set analysis, consistency is the percentage of cases in each row displaying the outcome

Parameters of fit are used to compare the various “causal recipes” that QCA identifies

• Consistency
  • # of cases displaying causal combination and the outcome / by the total # of cases displaying the causal combination
• Coverage
  • Refers to the number of cases with both the causal condition and the outcome divided by the number of cases with just the outcome.
• Proportional Reduction in Inconsistency (PRI)
  • Only applies to fuzzy-set analysis to deal w/ problem of simultaneous subset relations
  • Measure of the degree to which a given X is a subset of Y and not its negation

• Calibrate set membership
• Construct truth table
• Boolean reduction
• Compare causal recipes using parameters of fit

• Theory
• Visual Analysis
• Hierarchical cluster analysis (HCA)

• “Sect” refers to districts with pre-existing sectarian tensions
• Districts with the mixed presence of Sunnis and Shia/Alawi or homogenous districts from either group that border one another
• Eight districts of Lebanon’s twenty-six meet this definition
• Calibrated into 1’s and 0’s

• Parsimonious Solution
• Complex Solution
• Intermediate Solution
• All three will be examined (robustness)

 POPDEN + REF * SECT => CON

• The intersection of high refugee flows and sectarian tensions results in conflict
• In a majorty of cases with high population density, the outcome is conflict
• Though only few cases in our sample follow this path to the outcome

• Intermediate and Complex Solutions
• Resource Scarcity: Income < Population Density
• Fuzzy Set Analysis

POPDEN * REF * pov + popden * REF * SECT + POPDEN * SECT * pov => CON

• Low population density, high refugee flows, and sectarian tensions result in conflict
• Beirut is the only case to result in conflict without high refugee flows, likely bc of its status as the capital

REF * SECT + POPDEN * REF * pov + POPDEN * SECT * pov => CON 

• The intersection of high refugee flows and sectarian tensions results in conflict
• High population density, sectarian tensions, and low poverty result in conflict for a small # of cases

POPDEN + REF * SECT => CON

• The intersection of high refugee flows and sectarian tensions results in conflict
• There is less than consistent evidence high population density has an impact on coflict

• Combines set-theoretic analysis with gradations in set membership
• First introduced by Charles Ragin (2000)
• Relies on fuzzy set theory originally developed by Zadeh (1965)
  • Fuzzy sets are sets with a continuum of grades of membership
• Many social phenomena we study differ in both kind and degree
• Crisp sets calibrate memberships as a difference in kind
• But Fuzzy sets calibrate them as differences in kind and degree

Known as 'fuzzification'

• 1st step, identify the extreme poles of each condition
  • the threshold for full set exclusion
  • the threshold for full set inclusion
• 2nd step, identify the point at which a case belongs to both sets equally
  • the threshold for maximal point of ambiguity
• Rely on theory, visuals, and/or HCA to identify appropriate thresholds

 POPDEN*SECT*pov + REF*SECT*popden => CON

• Sectarian tensions are a necessary but insufficient condition for conflict
• Parsimonious and Intermediate solutions drop low poverty and low population density
• High population density and high refugee flows both work as interchangable demographic pressures that interact w/ sectarian tensions to result in conflict

• High refugee flows and sectarian tensions are a sufficient but not necessary causal recipe for conflict
• High population density is neither a sufficient nor necessary condition although a number of causal paths rely on high population density (far fewer rely on its negation)
• Income is either missing from most causal paths or produces inconsistent outcomes
• Overall, QCA provides strong support for EBP theory but less so for RS theory

• Introduced RS and EBP theories as conditional variants of refugees => conflict
• Exploratory analysis through GIS methods of hypervariate data visualization
• Introduced QCA method
• Applied QCA to the case of Lebanon: EBP theory holds firm, less evidence for RS
• Next step is to test EBP using spatial regression analysis
  • Extending temporal scope of data to the present
  • Including control variables