Network Analysis of Factors Contributing to Election Violence in Kenya: Comparing 2007-08 to 2013

Project Proposal: Peter Varnum and Mollie Zapata

This project will use networks analysis to assess election-related violence in Kenya in 2007-08, and then conduct the same analysis of 2013 when there was no election-related violence to determine if networks analysis can be used to predict violence. Additional factors in both cases will be studied separately: the networks of targeted violence in 2007, and how internal displacement in the interim period affected elections in 2013.

I. Overview: Kenya 2007-08

In December 2007, Kenya’s national elections led to two months of violent unrest across the country. In this politically motivated ethnic violence, police, youth militias, and community members killed at least 1,300 of their countrymen, which led to the displacement of an estimated 350,000 civilians. The international community was shocked. While ethnic violence was common and even expected in countries like Somalia, Rwanda, Sudan, and DR Congo, until 2007 Kenya had been seen as the beacon of stability in East Africa. 

Underlying this image of stability, however, were deep-seated ethnic, income inequality, land tenure, and other economic issues. Political parties in Kenya have always been unofficially formed along ethnic lines, with the Kikuyu-led Kenya African National Union (KANU) and Luo-led Orange Democratic Movement (ODM)  dominating. In response to what were deemed unfair elections at the end of 2007, Luo elders allegedly mobilized Luo youth militias to target Kikuyu living outside traditional Kikuyu areas. Kikuyus then retaliated, and the violence escalated across the country and continued through February 2008 when the leaders signed a power-sharing agreement.

Timeline of the Crisis:

• December 27 -- National elections begin.
• December 29 -- Opposition leader Raila Odinga of the ODM announces his victory.
• December 30 -- Incumbent President Mwai Kibaki of the Party of National Unity (PNU), a coalition of several parties formed in 2007 and dominated by the KANU, is hastily declared the winner by the Electoral Commission, despite voting irregularities and pre-election polls indicating that Odinga would prevail.
• December 30 -- Kenyan government bans public gatherings and live broadcast media (both illegal actions according to international law and Kenya’s Constitution).
• January 1  -- According to Human Rights Watch,“a mob set fire to a church where terrified Kikuyu residents were seeking refuge, soaking mattresses the victims had brought with them with petrol and stacking them against the building. At least 30 people were burned alive.” Note: This is just one example. Incidences of violence perpetrated by both sides escalated through January and continued through February. A 2010 report by the International Criminal Court noted that “crimes, such as murder, rape, mutilations, looting, destruction of property, arson and eviction seem to have occurred on the territory of the Republic of Kenya at least in the course of the events between 28/29 December 2007 and 28 February 2008.”
• February 5 -- International Criminal Court Prosecutor Luis Moreno Ocompo announces a preliminary examination of post-election violence.
• February 4 -- Government lifts ban on live broadcasts.
• February 8 -- Government lifts ban on public gatherings.
• February 28 -- President Kibaki and Raila Odinga sign a power-sharing agreement to end Kenya’s crisis.

Legacy of 2007 Elections

In the aftermath of the 2007 election, UN secretary-general Kofi Annan negotiated a power-sharing agreement between President Mwai Kibaki and Raila Odinga. The National Accord, as it was known, appointed Odinga the Prime Minister, and called for a referendum for a new Constitution of Kenya. The Constitution was passed in 2010 and divided Kenya into eight provinces -- the Rift Valley, Eastern, Northeastern, Coast, Central, Nyanza, and Western -- and 47 counties, whose size and boundaries are based on the 47 legally recognized Districts, or Sub-counties, of Kenya.

The continuation of violence for several months in early 2008 forced many people to flee their hometowns; the World Health Organization (WHO) estimates that 350,000 people were displaced. Though some internally displaced persons (IDPs) moved to IDP camps, many people -- particularly Luos -- migrated back to their ancestral homelands. As a result, the geographic division among tribes became more pronounced, with the Luos occupying the Rift Valley, Nyanza, and Western provinces and the Kikuyus occupying the others, in general. Despite this geopolitical homogeneity, little violence erupted after the 2013 elections.

Kenyan Elections 2013

The 2013 election, like the 2007 election, was controversial. Uhuru Kenyatta, the PNU incumbent president, who is Kikuyu, received 50.07% of the vote, according to the Independent Electoral and Boundaries Commission (IEBC); Raila Odinga, the ODM Prime Minister, who is Luo, received 43.28%. The two avoided a runoff, per the Kenyan Constitution (the Constitution dictates that one person must receive the majority of the vote to be elected President), by a difference of just 8,100 votes. Odinga quickly filed an appeal with the Supreme Court of Kenya, attesting that the IEBC counted votes fraudulently. The Court unanimously dismissed his case less than a month after the election.

Though the results were similar -- a close race between two rival ethnic parties, with allegations of voting irregularities from both sides -- very little violence occurred in 2013, unlike in 2007-08.

II. Project Plan

This project will determine what networks can tell us about election-related violence and whether networks analysis can be used as a tool to predict this type of violence. While extensive studies have been done on the causes and drivers of Kenya’s 2007 election violence, no one has yet used network analysis tools to assess how political and geographic organization contributed to unrest, nor compared that data to 2013, when the election results were similar but the aftermaths were completely different.

There will be two comparative analyses:

1. We plan to assess if and how the formation of political party networks contributed to election-related violence in 2007-08, and then, similarly, assess conditions in the 2013 elections to see if and how the networks changed.

2. We will also look at the networks of Kenyan communities based on election results in 2007-08 and 2013, with communities and the parties they support as the nodes and ethnicity as their primary attribute. We will see if patterns emerge and whether SNA could be used to indicate causality and possibly predict election-related violence.

Additionally, we will assess related factors for both election processes to gather a more complete picture of the situation.

2007-08: We will analyze who directed violence against whom over time to determine if the violence escalated due to a cycle of reprisals.

2013: We will assess if and how the internal displacement that was caused by the 2007-08 violence affected the way communities voted in 2013.

III. Objectives, Research Questions, and SNA Methods

Comparing 2007-08 to 2013:

Objective 1: To conduct a network analysis of inter-political party alliances in the 2007 elections to set a baseline to compare to a network analysis of inter-party alliances in 2013.

• Research Question: How did the network of party alliances change from 2007 to 2013?
• SNA Methods: We will analyze a one-mode dataset of political parties. This will be a dichotomized dataset with a “1” indicating that the parties are allied, and a “0” indicating that they were not.
• Attributes: 
• Number of members of each political party
• Ethnicity
• Geography of party base

Objective 2: To conduct a network analysis of how communities connected with (i.e. voted for) political parties.

• Research Questions: 
• Did the way county voted affect level of violence? 
• Were more homogenous counties less likely to experience violence? Alternately, were more diverse counties more likely to experience violence? Homogeneity will be assessed by voting records -- we can infer that those voting for a party are of the same ethnicity as that party, because political parties are overwhelmingly aligned with ethnicity.
• Did this change from 2007-2013? If so, how? 
• SNA Methods: We will start by analyzing bimodal data (counties and political parties), assessing the strength of ties based on percentage of votes (stronger ties = higher percentage of votes). This will be visualized by the width of ties between county nodes. We can infer a person’s ethnicity from the way they voted -- for example: a county voting 100% for the PNU can be assumed (for the purposes of our analysis) to be 100% Kikuyu. This will be a valued dataset, with the values corresponding with the percentage of votes going to each party. For example, if a county voted 20% for the PNU and 80% for the ODM, those values would be reflected in the weight of the ties.
• Attributes:  
• Incidences of violence reported in each community. (Node size will be larger for more incidences of violence.) 
• If possible, we will also consider attributes relating to other stresses present in the communities: Police force-perpetrated violence, land tenure issues, previous election-related violence, income inequalities, perception of misrepresentation in government.

Year-Specific Objectives:

Objective 3 (2007-08)*: To use networks analysis to assess patterns of violence in communities over time.

• Research Question: Who committed violence against whom? In what order? Can the networks tell us anything about why the violence escalated so quickly and continued for so long? 
• SNA Methods: We will select the counties that experienced the most violence in 2007-08 and use data of incidences of violence from Ushahidi and news reports to determine, over time, who committed violence against whom. This will be a bimodal dataset, with the nodes being: violence Kikuyu against Luo; violence Luo against Kikuyu; reciprocal violence; and no violence.  We will look at the directionality of ties. By doing a series of network analyses over time, we will assess whether the reason the violence escalated so quickly and completely was due to a cycle of reprisals that the government was unable to stop. 
• Attribute: 
• County

*Due to an inability to accurately determine who instigated violence against whom, it is unlikely that we (or anyone) will be able to conduct such an analysis for this project.

Objective 4 (2013): To use networks analysis to determine if and how internal displacement changed election results in districts from 2007-2013.

• Research Question: Could heterogeneity of communities cause election-related violence? If communities were not homogeneous in 2007-08 and they experienced violence -- and they were homogeneous in 2013 and did not (due to displacement/migration) -- then that could be an indicator that violence was caused by tribal heterogeneity within communities.
• SNA Methods: This will involve two datasets, combined. Each dataset will have ethnicity and counties as their axes. The first will answer the question, “Was there violence?” and yield either a 1, or a 0 (1 indicating there was; 0 indicating there was not). The second will answer the question, “Was there an influx of IDPs?” and yield a 10 or a 0 (10 indicating there were IDPs; 0 indicating there were not). We will then add the datasets together, resulting in a 0, 1, 10, or 11 meaning:
• 0: There was neither violence nor IDPs
• 1: There was violence, but no IDPs
• 10: There was no violence, but there were IDPs
• 11: There was both violence and IDPs.
• Attributes:
• Who perpetrated the violence -- civilians or police?
• How the county voted in the 2007 election.

Network Science at Center of Surveillance Dispute

Formulating counter-terrorism strategies for Pakistan using Social Network Analysis

Counter-terrorism strategies for Pakistan: Can Social Network Analysis identify targets for different strategies?

Arqam Lodhi for Social Networks in Organization Module-I. Unable to pursue this SNA proposal in Module-II.

Abstract: Different terrorist outfits and affiliates of Al-Qaeda network operating in Pakistan have wrecked havoc over the last decade, resulting in casualties of more than 25,000 civilians and security personnel since 2003. The newly elected government has been formulating a new comprehensive strategy, using a multi-pronged approach to deal with such outfits. Social Network Analysis (SNA) can be extremely valuable in identifying and isolating different outfits for targeted ‘treatment’, using all possible tools such as legislative reforms, negotiations and military operations. In this post, I have attempted to explore if SNA, done at both outfit and individual levels, can be used as a basis of counter-terrorism strategy. In the absence of a precise and customized approach, any strategy is unlikely to fully achieve its objectives.

Background:

Outfits’ Profile: According to some experts, it is hard to exactly determine how many unique outfits are operating in Pakistan. However, they can be categorized based on their declared objectives, as done by Ashley J Tellis of Carnegie Endowment for International Peace in a testimony before U.S. House Foreign Affairs subcommittee. These categories are: (1) Sectarian, (2) Anti-Indian, (3) Afghan Taliban, (4) Al-Qaeda and its affiliates, and (5) Pakistani Taliban. In addition to these, there are several other outfits such as Punjabi Taliban, separatist insurgent groups in Baluchistan province and some tribal elements in North Western Pakistan who either operate independently or in collusion with groups above in reaction to the changed demographic and political realities resulting from reduced autonomy after military operations on both sides of Pakistan-Afghanistan border. As of September 2013, Government of Pakistan listed and banned 52 organizations that are involved in terrorist activities.

Modus Operandi: Two elements are crucial for these outfits of any category: (1) money, and (2) manpower. In addition to these two, operational and logistical cooperation and ideology-based alliances connect these outfits. With funding sources from Middle Eastern diaspora and states drying up, most of these outfits also depend on an array of criminal activities such as drug trade, extortion, kidnapping, smuggling, money and trade laundering and looting.

Counter-terrorism strategy: The proposed counter-terrorism strategy should aim at dismantling the operational structure of these outfits, leaving them completely crippled rather than hoping to eliminate them entirely. The three potential approaches could be: (1) dismantling the financial resources pipeline and operational support structure, (2) negotiating with reconcilable elements, (3) targeted surgical strikes against outfits determined to be consistently operating on their agenda.

Social Network Analysis:

Objectives: SNA should be conducted at two levels, (1) organizational/outfit level, and (2) individual actor level. The key objective of SNA is to visualize how and what outfits are connected with each other; exploring the key determinants of these connections based on the attributes of these outfits and actors. A 2-mode analysis combining frequency of their interaction, based on the number of joint operations or communication, and their attributes would help us narrow down those outfits that have the strongest ties, i.e. connected by attributes, operational and financial cooperation and ideological position. Sub-group analysis should help us identify those factions within the network that might stand out in either extremes, willing or unwilling for negotiations. Evidence of dissent within militant outfits already exists and faction analysis would help us identify such outfits.

Key Research Question: Through network analysis, can we identify nodes and sub-groups within networks that can be targeted by three counterterrorism approaches stated above, based on their network position and attributes?

Network measures of interest: In Uncloaking Terrorist Networks, Valdis Krebs has identified three important network centrality measures to analyze these networks: (1) degreesàlevel of activity of a node within a network, (2) closenessàability to access others in the network, and (3) betweenessàability of a node to act as a broker, i.e. to control flow of information or resources between nodes.

Data: Following variables would be important to conduct this SNA. This is not an exhaustive list:

•  Total # of outfits (including same core/structure-multiple names)
• Level of interaction between outfits (communication, operational planning and execution, financial collaboration)
• Attributes:
•  Category 1 to 7 of outfit, listed above
•  Level of activism, determined by level of operational involvement/# of attacks
•  Attitude towards negotiation (state-led or mediated)
•  Ideological drivers: (political, religious, sectarian, ethnic, money or a combination)
•  Operational strength: (nature of attacks, # of members, weapons)
• Financial health

Limitations: Krebs describes three limitations to application of SNA to terrorist networks: (1) Incompleteness: difficulty to state with certainty that all actors have been accounted for in the network, (2) Fuzzy boundaries: challenge in deciding which actors to include in which category, and (3) Dynamic: these networks are always changing. Also, absence of ties or distance between nodes might not mean absence of connection or limited interaction; this might be done deliberately and strategically to deceive law enforcement agencies. Finally, data is very confidential and access to it is most likely restricted to officials in the security establishment.

Proposed SNA Project: Social Network Analysis of a School

Wah-Kwan Lin
Proposed SNA Project
I will be taking the second module of the Social Networks course for credit

Background
There exists a particular school with over 100 teachers and staff members, and enrolls over 300 students per year. In recent years, the school has undergone significant renovations and has introduced new leadership who are willing to experiment with innovations to improve both the educational and community experience.

Objective
The objective of this network analysis is to identify possible opportunities to modify the organizational structure of the teachers and staff at the school to enhance the learning experience for students and to strengthen the school community. The network analysis will ideally reveal possible opportunities for teachers and staff to collaborate with one another in manners that might not have been apparent before, and to reveal possible silos or bottlenecks that might be corrected to promote a more efficient and collaborative work environment.

Research Question
Can the organizational structure at a private school in Northern Massachusetts be modified to improve the educational and community experience?

Hypothesis
According to Cross et al. in the article, "Knowing What we Know: Supporting Knowledge Creation and Sharing in Social Networks", "who you know has a significant impact on what you come to know, as relationships are critical for obtaining information, solving problems and learning how to do your work." The ability to share knowledge can have very significant implications in knowledge and experience-intensive settings, such as schools. The informal social networks that arise in schools may be a significant factor behind the overall quality of the school.

Cross, in "Leading in a Connected World", also suggests that network information can be leveraged to significantly improve the overall success of an organization by identifying bottlenecks, encouraging key actors, better integrating peripheral members of the organization, engaging high performers, and applying individuals with appropriate skill sets to unique challenges. Network information can potentially be used to similarly promote success within a school setting.

Kreb, in "Managing Core Competencies of the Corporation" additionally suggests that the actual networks that arise may differ substantially from the officially established hierarchical structures. Network analysis may reveal surprising realities about how information flows, how individuals and groups interact, and whether or not individuals or groups are sub-optimally positioned within a larger organization. A network analysis of a school might similarly reveal differences between the intended hierarchical structure and the actual structure that operates in practice.


Methodology
Currently, the data that would be appropriate to conducting a network analysis of the school in question is not available. A survey would therefore have to be conducted. For logistical and pragmatic reasons, the survey will only focus on adult employees of the school, and it will not focus on the teenaged students of the school.

Possible questions to be asked on the survey include:

• Are you a teacher or a member of the staff/administration?
• How many years have you worked at the school?
• What is your primary mode of communicating with colleagues? Face-to-face? Phone? Email? Instant Messaging? Text messaging? Other?
• On average, how many students meet with you outside of class per week?
• Who do you turn to for teaching advice or support most frequently?
• Who do you turn to for non-teaching related advice or support most frequently?
• Who has shared lesson plans or other collections of knowledge with you?
• Do you feel that you have too many students in your classes? Too few? Just enough?
• If you could teach any class or pursue any project that does not currently exist, what would it be?

Once the survey data is collected, it can be analysed for:

• Centrality, which might reveal whether or not individuals are well integrated into the school community.
• Indegrees, which might reveal key individuals in the school that are heavily relied upon by other teachers or staff. These might represent stars in the school, but they might also be bottleneck points.
• Outdegrees, which might reveal individuals who may require additional support and resources.
• Eigenvector centrality, which might reveal individuals who are key points of knowledge dissemination in the school.
• Cliques, which might reveal closely collaborative groups among the teachers or staff, or it might suggest the possibility of groupthink.

Additional Considerations
Whether or not this project can proceed is contingent on the school's approval.

As this analysis will be based on survey data, typical survey-related challenges should be expected, including incomplete or complete lack of responses from potential participants and inaccuracy or bias in the responses.

Legislative Voting Patterns and Party Defection in the Japanese National Diet

Ryo C. Kato

(will not be doing this in 2^nd module)

Background

In 2009, the Liberal Democratic Party (LDP), which had been the governing party since 1955 (except for 11 months in 1993), was handed the worst defeat of a sitting government in the history of Japanese electoral politics. The opposition party, the Democratic Party of Japan (DPJ), captured 308 seats out of 480 in a landslide, and secured the office of Prime Minister. This event was touted as a major landmark for Japanese democracy and was interpreted as a repudiation of the LDP for its domination over post-war politics. However, after the 2012 elections, only three years after utter defeat, the LDP was back in power. In both elections, there were a number of legislators, some of who were instrumental to the outcome of the elections, who had defected from their parties. The kingmaker of the DPJ, Ozawa Ichiro, for instance, was a powerful chief secretary of the LDP before defecting in 1993 to form a series of opposition parties with other former members of the LDP, one of which was the DPJ. In 2012, he once again defected and formed a new party, the Life Party, taking with him several former DPJ members.

A recent paper showed that there is a correlation between seniority within a party and switching affiliations. It found that very junior or very senior legislators were more likely to switch. What are other characteristics of the politicians who switch party affiliations? Do they hail from the same prefectures, do they represent similar industrial and commercial interests in their constituencies, do they sit in the same committees, have they spent time in the same or similar companies, or are they old-boys of the same universities? Or is it voting patterns that correlates most with party switching?

Primary Question

Can patterns of voting on a selection of bills be used to predict the cohesiveness of political parties in Japan? Restated: when legislators switch political parties or create new ones, do they do so with legislators with whom they had voted together on important bills?

Hypothesis

How politicians switch their party affiliations correlates with patterns of voting on legislation and connection to powerful kingmaker politicians.  

Data

I will need two sets of networks for the members of the National Diet, one from immediately before and one from immediately after the 2012 elections. Current legislators are listed on websites maintained by the two-houses, the House of Representatives and House of Councilors. Although historical compositions of the legislature are not kept on these websites, it is relatively easy to find this information from government databases and academic sources online. While the information of specific legislation at various points in the law making process are easy enough to find, the record on how legislators voted on them is a different matter. While the House of Councilors website publishes their members’ voting records on bills, the House of Representatives’ website does not. From what I have gathered, finding voting records on bills introduced to the House of Representatives will require requests for data from a government office. 

Attributes for the legislators will include, party affiliation, age, prefecture where their district is located, alma mater, category of industry they worked in prior to political life,

I will manipulate the two 2-mode network of legislators and bills, one each for before and after the 2012 elections, into a pair of 1-mode networks of legislators, where the ties represent common votes on legislation. The thickness of the line will represent the number of common votes.

Important Network Measures

A large part of this assessment will follow looking for visual patterns and comparing the pre- and post-election networks. The first, the 2-mode network of legislators and legislation, will look for clusters of nodes that vote similarly on legislation. There will be on visualization showing edges representing ‘nay votes,’ and another where the edges are ‘yay votes.’ I will compare the clusters of the pre- and post-election networks.

The second visual will be used to look for patterns in 2-to-1 mode networks of the pre- and post-election. The nodes represent legislators and the edges represent shared votes, and the thickness will show the number of shared votes. Do clusters in the pre-election networks resemble clusters in the post-election network? Do clusters based on party affiliation in the post-election network resemble clusters in the pre-election network based on other attributes (shared votes, shared committees, etc).

I would also look at measures of degree centrality in the pre- and post-election 2-to-1 mode networks of legislators with ties indicating shared votes. If shared votes do in fact correlate with party affiliation, then power politicians that carry other legislators to other parties, such as Ozawa may exhibit high degree centrality in this network.

Conclusion

Social network analysis can help narrow the study of Japanese legislator’s switching political affiliations in the National Diet. What are other attributes other than age that correlate with this behavior? Do similar voting patterns between legislators in a party or in the factions within that party correlate with party defection? Or are the connections that share many attributes with influential politicians, such as Ozawa, more correlated with party defection?

Additionally…

Given the historical strength of certain political families, some of which trace their lineage to the leaders of the Meiji Revolution, it may be of interest to conduct an intertemporal and intergenerational examination of political cooperation and party cohesion. Are the network characteristics of the Meiji oligarchs or post-war political elites reflected by their grandsons and great-grandsons?

Proposed SNA: Social Network Analysis as a means of impeding networks of exploitation among Tajikistani labor migrants

Introduction/Background:
I spent the last summer working for American Councils on International Education in Dushanbe, Tajikistan, as director of the Persian Critical Language Scholarship program. While working there I also conducted research on language policy in Tajikistan as it relates to labor migration.

Labor migration is an extremely important topic for Tajikistan. The country has always been the poorest in Central Asia, but the Tajikistan Civil War from 1992 to 1997 further crippled the economy and its lingering effects have continued to hamper growth and economic opportunity for Tajikistani citizens. This encourages many Tajikistanis to seek employment abroad and it is estimated that 90% of Tajikistani migrant workers find employment in Russia. Migrants choose Russia because of its visa-free regime for Tajikistani citizens (non-employment seeking visits are permitted up to 90 days), the shared history between the two countries as members of the Soviet Union, and the presence of migrant networks and communities that have existed since the time of the USSR.

Since the fall of the Soviet Union, however, knowledge of the Russian language has become less widespread in Tajikistan. Many Russian speakers left during the civil war and the education system is so impoverished that it has proved capable neither of providing necessary materials nor of implementing pedagogy to teach Russian effectively in the country’s schools. This lack of Russian language competency is a big disadvantage to the newest generation of labor migrants. It complicates their efforts to find employment in Russia and makes it more difficult for them to understand their rights and responsibilities under Russian law. This exposes them to exploitation by Russian employers and their intermediaries.

In previous social network studies, researchers (Phillips and Massey 2000)(Waldinger and Litcher 2003) have shown how social networks have been able to help migrants in the moving process and in finding jobs. Most studies performed have highlighted these positive “social capital” aspects of migrant social networks. In her paper entitled “Networks of Exploitation: Immigrant Labor and the Restructuring of the Los Angeles Janitorial Industry”, however, Dr. Cynthia Cranford showed how existing social networks could lead to negative consequences for immigrants. Her research with janitors showed that immigrants were usually exploited by intermediaries (often the same ethnic group as the workers they exploited) who were in turn put under extreme stress and pressure by an American company to provide cheap labor. Cranford found that workers with strong ties to supervisors (family member or close friend) experienced less exploitation while those with weak ties (heard about the job from a neighbor or distant relative) were more likely to be exploited.

I would like to design an SNA that would trace networks of exploitation among Tajik workers in the construction industry.

Goal: Identify exploitative avenues to finding work in Russia so that Tajik migrants can be steered towards safer alternatives. 

Questions and Methodology: Tajik migrant workers in the construction industry in Moscow, Russia would be asked who the most influential person (people) have been in helping them to find their current job in Moscow. They would also need to be asked a series of questions about their employment experience to determine if it has been exploitative. The exploitative or non-exploitative nature of the job will be recorded and added to the attribute file along with information such as their gender, age, a self-assessment of their knowledge of Russian, the region of Tajikistan they hail from, etc.

The hope is that a network would emerge of both positive and negative (exploitative) sources of information regarding employment for Tajik workers in Moscow. This information could then be communicated to newer waves of migrants as they plan their entry to Russia.  

U.S. Political Polarization Over Time

I'm sure you've all seen this already on Facebook, but here is a visualization showing relative difference and similarities between Republican and Democrat senators voting over time.

http://today.duke.edu/2013/05/us-political-polarization-charted-new-study#video

A live map of DDoS attacks

http://www.digitalattackmap.com/#anim=1&color=0&country=RU&time=15999&view=map

Datasets for social network analysis

Still searching for datasets that you could analyze for a potential social networks project? Here are some resources I've found:

• Yahoo! Labs - http://webscope.sandbox.yahoo.com/catalog.php?datatype=g
• Internet Topology Zoo - http://topology-zoo.org/
• Trust Network Datasets - http://www.trustlet.org/wiki/Trust_network_datasets#Released_datasets
• BitTorrent User Activity traces - http://cs.brown.edu/~pavlo/torrent/
• Correlates of War - http://www.correlatesofwar.org/
• Terrorist Networks - http://www.orgnet.com/hijackers.html
• Combating Terrorism Center Harmony Network (the data is very raw... you'll find things like handwritten reimbursement requests by al-Qaeda operatives, profiles of suicide bombers in Arabic, etc) - http://www.ctc.usma.edu/programs-resources/harmony-program
• Teaching Terrorism - http://www.teachingterror.net/data.htm
• InfoChimps - http://www.infochimps.com/tags/social-networks
• CAIDA Internet Data Analysis - http://www.caida.org/data/overview/
• Quora listing of datasets - http://www.quora.com/Data/Where-can-I-find-large-datasets-open-to-the-public
• Dynamics Lab - http://www.growmeme.com/sna/visual
• Databib - http://databib.org/
• Datahub - http://datahub.io/
• Reddit datasets (Reddit proves its worth once again!) - http://www.reddit.com/r/datasets
• R data mining - http://www.r-bloggers.com/datasets-to-practice-your-data-mining/
• Youtube Dataset - http://netsg.cs.sfu.ca/youtubedata/

Good luck!

Another article on Chicago's gang violence and social networks

"Social-network policing is the future of crime fighting."

http://www.govtech.com/public-safety/Social-Media-Transforms-the-Way-Chicago-Fights-Gang-Violence.html

What Is Net Neutrality, and Why the War of Words?

Should a handful of large private companies be able to charge some businesses more or less for access to broadband Internet? Verizon, Time Warner, AT&T, and Comcast think so.


http://www.minyanville.com/sectors/technology/articles/What-Is-Net-Neutrality-and-Why/9/10/2013/id/51677#ixzz2huOgvigD

SNA in the NSA

On my drive home from Fletcher this evening, I caught the tail-end of a NPR segment about the NSA's use of meta-data and SNA. I couldn't help but smile as I realized I actually understood what they were talking about. As one would expect, the NSA is using SNA to track emails, trace associated contacts, and let SNA "do its thing". Here's a similar story posted on NPR.com.

*As a side point: while "SNA in the NSA" does have a certain musical ring to it, members of our seminar will be quick to tell you that the in-class "BM redundant of HA" SNA rap was one for the record books. Thanks Professor Tunnard.*

Freedom on the Net 2013

For those interested in Internet freedom by country, here's Freedom House's annual report for 2013.

Five Trends Shaping the Future of Work

http://www.forbes.com/sites/jacobmorgan/2013/06/20/five-trends-shaping-the-future-of-work/

4 Social Media Tools to Keep Your Office Connected

http://www.psfk.com/2013/02/social-media-tools-office.html

Open Internet = More Secure Internet?

This is debatable!

University systems overloaded with wireless devices

This isn't exactly about the inequality of IP number distribution, but it points to the same problem: there is an explosion of devices that access the Internet, and it's expensive for WiFi providers (like universities) to keep up with demand.

Electronic Frontier Foundation Resigns from Global Network Initiative

This is a pretty big deal. The EFF and GNI have been partners in combating attempts to rein in Internet freedom, but the EFF has pulled out, citing "a fundamental breakdown in confidence that (GNI's) corporate members are able to speak freely about their own internal privacy and security systems in the wake of the National Security Agency (NSA) surveillance revelations." Follow the link to the GNI site and download their 2012 annual report. Fodder for the debate...

Homophily, Cliques, and Organizational Performance

This is an interesting paper, relevant to what we've been discussing in class recently. Link to the article is in post tile, but here's an extract:

In a Working Paper entitled “Does Homophily Affect Performance?” Gargiulo and Assistant Professor of Strategic Management at Singapore Management University, Gokhan Ertug (INSEAD Ph.D. in Management, 2008) focused on nationality as a shared characteristic, and how it affects job performance on investment bankers. “If you are new, you’ve just been hired at an investment bank or consulting firm, it’s a difficult environment, you are likely to look for people who could be more helpful,” Gargiulo explains. “You don’t know it, but you would expect people of the same nationality as you, especially if you’re a minority, will be more naturally inclined to help. That’s why we chose nationality.” Gargiulo continues, “We chose investment bankers because we wanted a context in which informal relationships of knowledge transfers are very consequential for the performance of the individual. I mean, this is true of any organisation today, but in this knowledge-based organisation such as investment banking and consulting, it is particularly the case.”

Show me the money! The paper studied 1,746 investment bankers at a major international bank, with their performance measured by the bonuses they earned. These bankers were classified into four ranks, from most junior to most senior: Associate Director; Director (Vice-President); Executive Director (Senior Vice-President); and Managing Director. After controlling for factors that might affect how much bonus a banker earns e.g. tenure, rank, age, function, etc, Gargiulo came to one main conclusion: homophily helps a new hire, but hinders an experienced banker gunning for a promotion into senior management. (Emphasis added.)

Hashtag analysis of Azerbaijani elections tweets

Today Azerbaijan, a small oil-producing country on the Caspian shores, held their presidential elections. Ilham Aliyev, sitting president since 2003, when he took over after his father, won with a great majority of the votes. (The result was actually leaked one day before the elections, so that was no big surprise!) However, what is interesting, is this analysis of the opposition and pro-regime youth groups' use of twitter hashtags in the run-up before the elections. Katy Pearce has been logging the main hashtags connected to the elections for more than a month, and finds that twitter has become an important battlefield between opposition groups and pro-regime youth groups before the elections. Thus, while the oppression of human rights and the opposition in particular in Azerbaijan continues, the regime still seem to care about their reputation...

http://www.katypearce.net/we-are-young-heartache-to-heartache-we-stand-no-promises-no-demands-azvote13/

"It takes a network to defeat a network"

Really interesting article on the comparison between insurgencies in Iraq and Afghanistan and how social network analysis came into play in the evolution and execution of COIN.

http://insna.org/PDF/Connections/v33/Knoke_Vol33Iss1_INSNApdf.pdf

Data Sources for Conflict Analysis

I have come across three excellent starting points for those looking for armed conflict-related data:

“A Beginner's Guide to Conflict Data: Finding and Using the Right Dataset" by Kristine Eck was published by the Uppsala Conflict Research Program and is a goldmine for anyone looking for conflict-related data. The first part of the paper gives an overview of factors to consider in data research, and the second part is a list of 60 conflict datasets with descriptions, source/accessing information, and sorted by armed conflicts and events.  

The Correlates of War Project (or COW) was established with the goal of “the systematic accumulation of scientific knowledge about war.” The COW database includes wars from 1816 to 2007. COW defines international war as “a conflict waged between (or among) national entities, at least one of which is a state, which results in at least 1000 battle deaths of military personnel.” The project also features data sets of World Religions, Alliances, Trade, Diplomatic Exchange, Intra- and Inter-state war. Under the “Related Data” tab, the project links to many other potentially useful data sets.

The Uppsala Conflict Data Program (UCDP) collaborated with the Center for the Study of Civil War (CSCW) to create the UCDP/PRIO Armed Conflict Dataset, which differs from the COW project in that it defines a conflict based on 25 or more battle deaths. This data set includes conflicts from 1946-2013. Notable other data sets include a Geo-referenced Conflict Site dataset, and the Armed Conflict Location and Event Data, which codes exact locations, dates, and additional characteristics of individual battle events in states affected with civil war.

Using News Sources for Social Networks Analysis: The Sudan Tribune from 2003–2010

In a paper published in 2012, authors use analysis of keywords in online news articles, data on ethnic group distribution, and geospatial information on the locations of ethnic groups and conflicts to correlate ethnic violence with key words. The paper, "Structure of ethnic violence in Sudan: a semi-automated network analysis of online news (2003–2010)" assesses the potential for conflict in Sudan before the secession of South Sudan.

The authors tested "whether an ethnic group's connections to the environment (livestock, biomes, and other resources) and other ethnic groups was associated with severe conflict and peace terms and whether ethnic-group richness at a given geospatial location was associated with severe conflict."

One of their findings was, "that resource linkages among ethnic groups were a determinant of severe conflict. Those groups that had either a strong connection to one biome or those that were tied to multiple biomes had more peace. This suggests that competition over the means of livelihood is critical. The finding that overlapping ethnic groups also is tied to conflict also lends supports that conflict may be over competition for resources."

"The first sphere of influence was the ethnic-group organization by environmental-resources network (Fig. 2a) and a second sphere of influence was the ethnic-group organization by socio-political knowledge network (Fig. 2b)...For example, the Murle ethnic group environmental-resource sphere of influence network (Fig. 2a) shows that the Murle were tied to the Dinka, Bor, and Nuer ethnic groups via livestock....RESULTS: An ethnic group that had more ties to CONFLICT within their sphere of influence also had more ties to LIVESTOCK."

The authors did statistical analysis in addition to social networks analysis. Full paper available: http://www.tracyvanholt.com/pdfs/vanholt_etal_CMOT_2012.pdf

How many people use the top 275 Social Media sites

As comprehensive a list as I've found. All stats are linked to their source sites, so you can decide for yourself if the measurement is independent, or company-hyped.

World's most popular websites--by country (uses Alexa data)

The network that was behind the government shutdown? There's a potential project in here.

An article in today's New York Times refers to this letter (click on title) as the "blueprint" for the resistance that led to the standoff in Congress. If you start with the signatories, you might have the beginnings of an interesting network project... (Added Monday, October 7th: Another NYT article on Cruz, the Republican's use of Social Media, and the shutdown here)

"E-mails tend to flow much more frequently between countries with certain economic and cultural similarities."

While exploring possible social network projects, I came across this little nugget: http://www.washingtonpost.com/blogs/worldviews/wp/2013/03/07/an-incredible-map-of-which-countries-email-each-other-and-why/

Basically, it's a social network of countries, with the strength of the ties between countries reflected by the volume of emails between the countries. The findings suggest that that analysis provides support (though qualified) for Samuel Huntington's "Clash of Civilizations" argument.

A Networked Model for Drug Trafficking in Mexico

"In this model, trackers minimize the costs of transporting drugs from producing municipalities in Mexico across the road network to U.S. points of entry. They incur costs from the physical distance traversed and from crackdowns and thus take the shortest route to the U.S. that avoids municipalities with crackdowns."

Really exciting stuff from a brilliant post-grad:

http://scholar.harvard.edu/files/dell/files/121113draft_0.pdf

Documents on N.S.A. Efforts to Diagram Social Networks of U.S. Citizens

New documents leaked by Edward J. Snowden, the former N.S.A. contractor, provide a rare window into the N.S.A.’s push to exploit phone and e-mail data of Americans after it lifted restrictions in 2010.

Suggested impacts of the approach:

" In the first place it allows NSA to discover and track connections between foreign intelligence targets and possible 2nd Party or US communicants. In the second place it enables large-scale graph analysis on very large sets of communications metadata without having to check foreignness of every node or address in the graph."

http://www.nytimes.com/interactive/2013/09/29/us/documents-on-nsa-efforts-to-diagram-social-networks-of-us-citizens.html

Using social media to get hired

New research from Tufts economist Laura Gee, questioning the power of Granovetter's weak ties when it comes to landing a job:

http://www.forbes.com/sites/adamtanner/2013/10/03/the-odds-of-facebook-telling-us-anything-about-how-to-get-a-job-are-iffy

Vikings: burly savages or social network sages?

New research sheds light on a "complex social network at play" in the times of Vikings:

http://www.natureworldnews.com/articles/4270/20131001/viking-society-contained-vast-social-networks.htm

Big Data and Big networks--in Boston

This will be a really interesting event for those interested in--well, have a look. And you can go after the debates!

"How Social Networks Explain Violence in Chicago"

Video about how risk factors alone do not predict violence, but how relationships are key. Via The Atlantic, whose October issues explores SNA and violence in greater depth.

Social networks in innovation diffusion

Interesting article about how social network analysis leads to focus on certain networks that work in technology diffusion (in this case, family networks) as opposed to others that don't (religious networks).

Social Networks and Technology Adoption in Northern Mozambique

http://econ.lse.ac.uk/staff/bandiera/technology.pdf

Tracking the "the most powerful operative in the Middle East"

The New Yorker article referenced in class today that looks at Qassem Suleimani, the head of Iran's Quds Force who is directing the war in Syria for the Assad regime. Of particular interest for our SNA class is the tool of "co-location analysis" which prosecutors used to identify the members of Hezbollah allegedly involved in the assassination of former Lebanese PM Rafik Hariri. Investigators tracked calls made on disposable phones to phones that were known to be owned by the suspects at the time of the assassination. A long read but totally worth it...

http://www.newyorker.com/reporting/2013/09/30/130930fa_fact_filkins?currentPage=all

Network Analysis in Macroeconomics and Development Literature

The figure below shows the network of countries based on the similarities in exports (or Product Space). Key finding of the paper: Countries are technologically similar to their neighbors.

'The figure above presents the network of export similarity for year 2008 as a graphical network where each node represent a country, and each country is connected to the two other countries with the most similar export baskets, as measured by the Export Similarity Index. Countries are colored according to geographic regions, showing that the clusters defined by export similarity correlate strongly with physical distance. The width of the links is proportional to the similarity index and the color of the link indicates whether the similarity is driven by PRB products (blue) or by NPRB products (red) (see link for more details). We note that, in a large number of cases, the country with the most similar export structure is an immediate neighbor, such as in the case of France, Germany, Austria, the Czech Republic, Hungary and Slovakia or in the case of India, Pakistan, and Bangladesh. This visualization illustrates the strong association between proximity and export structure that characterizes the world economy.'

Check out the paper at http://dash.harvard.edu/handle/1/8830781