Friday, October 25, 2019

Do board interlocking networks influence high performance in corporate sustainability?

I would like to explore the following question:

Do board interlocking networks influence high performance in corporate sustainability?

Board interlocking networks are complex social networks among company boards created by directors sitting on multiple boards.[i] Many prior studies have examined the origins and effects of board interlocking networks in different contexts and across various time periods, showing that firms’ actions are often related to their interlocks. For example, in a recent study, Wong, Gygax and Wang (2015) present evidence that board interlocks are positively linked with similarities in the design of executive compensation packages in interlocked firms as well as in a number of board characteristics.[ii] In the context of corporate strategy, Davis (1991) demonstrated that there is an interlock network diffusion process with regard to the use of poison pills as a defense strategy.[iii] Haunschild (1993) showed that corporate acquisition activities are spread over the interlocking network by imitation processes.[iv] Palmer et al. (1993) demonstrated that interlocks are associated with the adoption of a multidivisional form within corporations,[v] and Galaskiewicz and Wasserman (1989) showed that interlocks are linked with how corporations decide on their charitable recipients.[vi]

By sitting on multiple company boards, directors can facilitate the sharing of information, experience, and other techniques and managerial practices. I would like to investigate whether there is evidence that robust sustainability management practices are diffused through board interlock networks. My hypothesis is that the existence of a board interlock between two boards is correlated with the similarity in corporate sustainability strategy, governance and performance in the two companies. In other words, I expect to observe similar corporate sustainability approaches and outcomes in connected firms.

In order to test this hypothesis, I propose to conduct a social network analysis of the board interlock network of a selection of companies included on the Corporate Knights Global 100 (CKG100)[vii] list of the most sustainable companies in the world (published annually from 2005 to 2019) at three points in time – 2009, 2014, and 2019. Each year’s network would include the selection of CKG100 companies as well as the firms with which they are interlocked by one or more board members (this data may be extracted from a database such as BoardEx, GMI Ratings Corporate Governance Database, or ORBIS database).

For each of these networks, I would then examine the homophilia (E-I index) of companies related to the following attributes: CKG100 inclusion/exclusion status, existence of a board level sustainability committee, and existence of an executive level sustainability position within the firm (this data should be available in companies’ corporate sustainability reports). I would also conduct various sub-group analyses (clique and newman-girvan) to determine if any sub-groups tend to form on the basis of the same attributes. Finally, I would analyze the evolution the networks over time to see if there was an increase in homophilia and/or sub-group formation on the basis of these attributes over time, indicating an influence of board interlock networks on corporate sustainability strategy, governance and performance over time.

[i] Mizruchi, M.S., 1996. What do interlocks do? An analysis, critique, and assessment of research on interlocking directorates. Annu. Rev. Sociol. 22, 271–298.
[ii] Wong, L., Gygax, A., Wang, P., 2015. Board interlocking network and the design of executive compensation packages. Social Networks 41, 85-100.
[iii] Davis, G.F., 1991. Agents without principles? The spread of the poison pill through the intercorporate network. Adm. Sci. Q. 36, 583–613.
[iv] Haunschild, P.R., 1993. Interorganizational imitation: the impact of interlocks on corporate acquisition activity. Adm. Sci. Q. 38, 564–592.
[v] Palmer, D.A., Jennings, P.D., Zhou, X., 1993. Late adoption of the multidivisional form by large U.S. corporations: institutional, political and economic accounts. Adm.Sci. Q. 38, 100–131.
[vi] Galaskiewicz, J., Wasserman, S., 1989. Mimetic processes within an interorganization field: an empirical test. Adm. Sci. Q. 34, 454–479.
[vii] The Corporate Knights Global 100 is one of the most widely publicized and well-respected corporate sustainability rankings, based on companies’ public disclosure of 21 key performance indicators (KPIs) covering resource management, employee management, financial management, clean revenue and supplier performance.

Tuesday, October 22, 2019

Effect of Trade and Environmental Treaty Networks on Trade in Environmental Goods & Services

In the light of the failure of the Doha Development Round of the World Trade Organisation (WTO), preferential trade agreements (PTAs) have proliferated in the last 25 years. From around 50 PTAs in force in 1991, this number has grown to 481 in 2019. The PTAs now form a complex network as shown below:

PTAs have evolved from “shallow” agreements that cover tariff and non-tariff barriers to “deep” agreements that cover a host of other issues like labour, environment and intellectual property rights. [1]

Many PTAs now cover environmental aspects such as trade in environmental goods and services, and compliance with multilateral environmental agreements to which the signatories of the PTA are already a party. Bilateral and Multilateral Environmental Agreements (BEAs and MEAs) also form a complex and overlapping network and may include trade in environmental goods and services.
However, traditionally growth in trade and environmental protection have been seen as opposing forces rather than complementary goals. Are the most central actors in the trade network also providing leadership in trade in environmental goods?

1)     Which network has greater effect in increasing trade in environmental goods and services – PTAs or BEAs?
2)     Are the most central actors in trade networks also providing leadership in trade of environmental goods and services? Or can we identify new actors in this space?

Data will be used from the following sites:
2)     RTA Exchange:
3)     Database of MEAs:
4)     World Bank Database of World Trade statistics:
5)     WTO Environmental Database:

The network of PTAs as well as the network of MEAs are two-mode, nondirectional networks, which may be given a value as deep or shallow agreement. This will be correlated with the network of actual flows of environmental goods and services across borders, which is a one-mode, directional network, valued by amount of trade (low =1, medium =2, high = 3).

Centrality measures will help us see whether the same actors are leaders in signing both PTAs and MEAs and whether this translates into actual leadership in trade in environmental goods. Here, degree centrality is not of as much use as signing of treaties is no guarantee of actual increase in trade. A country could sign a large number of treaties with actors that are peripheral in the international trades network. Hence, Eigenvector centrality will be given more importance. A large betweenness score could indicate that the country acts as an important trade centre in connecting two otherwise loosely connected economies.

Density & Centralisation
Density will help us understand the potential linkages while Centralisation will identify hubs in the trade network.

Subgroup Analysis
It will be useful to identify cliques in the trade of environmental goods and whether flows are predominantly one of the four:
1)     North-North
2)     North-South
3)     South-North
4)     South-South

My hypothesis is the first two types of flows dominate when it comes to trade in environmental goods and services. Subgroup analysis will help us identify the hubs and spokes of the system or identify a core-periphery structure if it exists, in trade in environmental goods and services.

Social Network Analysis is a useful tool here besides traditional economic analysis because it can help identify a core-periphery structure in the trade of environmental goods through centrality, whole network and subgroup analysis. It can also help highlight potential linkages in the network and identify new opportunities and potential leaders/hegemons in trading environmental goods.

If I take the second module, I will only be restricting my analysis to Eurasia due to shortage of time.


Monday, October 21, 2019

Mapping the Radicalization of Incels and the Spread of the Incel Network


There is a rising movement of people who identify as “incels,” –shorthand for “involuntary celibates.” This is a growing group of people (primarily men) who believe that they are oppressed by society’s strict status quos and norms that effectively prohibit the ugly or socially awkward from having sex. What began as a type of online support group for the dateless morphed into forums for misogyny and hate. With some veins of influence from the alt-right and sexism, incel networks are a growing danger, specifically for women. The extremist views of incels are often developed over time, and online networks use radicalization techniques not unlike those of ISIS and other extremist groups to gradually groom men to ascribe to extreme views about women. More and more, the hate speech and online vitriol is translating into physical violence. There have been several terrorist attacks against women, like the 2014 University of California Santa Barbara killing spree[1], a Toronto van attack[2], and the 2018 gunning of a Tallahassee yoga studio[3].

There is fairly little research on incel networks to date, with merely a few Vox articles sparking interest in the public. The ascension of the network is interesting, because the original founder of the incel movement lost control of the narrative long ago and the network morphed from support for people who felt ostracized into an echo chamber that blamed women, ironically the object of the incels’ desires, for their inability to connect with them. There is a great deal of crossover between incels and fringe-right forums like 4chan, which may lend some insight into how extreme ideas develop and how incels have broadened their network and reach so quickly.

Research Questions:

There are several research questions that I hope to address using Social Network Analysis to study this topic:
  • Who are the current leaders/voices in the incel communities?
  • Are there sub-communities (cliques) within broader incel networks? What types of connections do they have with the leaders of the broader communities?
  • Are there incel cliques within other types of extremist networks (like alt-right networks)?
  • How have incel networks grown over the past ~20 years?
  • What other networks do they overlap with? Are incels actively recruiting from these other networks or is there a common set of interests that naturally pulls people in from alt-right networks?
  • How are incels radicalized? How long does it take for incels to become radicalized? What type of media garners the highest amount of influence/interaction?


These Reddit-based networks will be most important, including: alt-right networks, “manosphere” networks, white supremacy networks. Since we tend to know more about alt-right and white supremacy networks, betweenness and degree measures between these and incel networks will be crucial to understanding the spread of extremist messaging and the radicalization of incels. Centralization will be particularly critical to learning how many sources of information there are in the networks. This will hopefully help sites with content moderation as well as provide valuable information for law enforcement. Countering the narratives of incel leaders is important, but it will be crucial to figure out who these leaders are first.

Data and Methodology:

While there is not extensive data on incels, their forums do collect data in informal polls. This Vox article[4] cites the Braincels forum’s poll data as well as other forums’ polling data, which I believe I can obtain by contacting the author Zack Beauchamp. Another preexisting dataset is the  Southern Poverty Law Center’s Hatewatch group’s logs from an Discord server, an online chat platform favored by gamers, from the day after the Toronto attack.[5]

Beyond this, I plan to resort to scraping social media data, primarily from Reddit, one of the most highly-used platforms for incel networks. By determining which users span across forums, as well as how many posts the users make, we will be able to target the sources of influence. This combined with qualitative research on the potency of the posts (as measured by slurs, hate speech, allusions to violence, scaling all the way up to direct threats) will lend some insight into how escalation and radicalization happen over time.

Additionally, scraping data for references praising known incel terrorists who have committed terror crimes such as Elliot Rodger, Alek Minassian, and Scott Beierle will expose a baseline of terrorist sympathizers.

Why Social Network Analysis:

Social Network Analysis is the best method to use to answer these questions because it is crucial to trace the spread and reach of information through terrorist networks and learn how to break up the network. SNA will allow me to observe the overlap between networks and see if any insights can be gained by observing the content generated by the forums. Online forums do present new obstacles, however, because it is difficult if not impossible to trace the connections or directionality within a forum. One potential workaround is to assign directionality based on likes and comments / responses made to an original post.

Using SNA to model connections between, capabilities of, and common indicators among fighting forces in Syria

Research Question:
 àWhat can modeling fighter groups and their affiliates reveal about alliances in the Syrian conflict?

Why Social Network Analysis?
àThere has been much written about the eight-year long Syrian Civil War, particularly given recent operational decisions taken by U.S. President Donald Trump and Turkish President Erdogan. However, much of the media coverage fails to fully consider the chaotic mix of fighting forces on the ground in Syria. Pro-government forces include the Syrian army and a number of affiliated Shia militias, Russian forces and private military companies such as the Wagner Group[1], Iranian Revolutionary Guard Corps forces[2] and several allied proxy organizations including but not limited to Hezbollah. Opposition forces include the Free Syrian Army (FSA), the formerly U.S.-backed Syrian Democratic Forces (SDF) which are comprised mostly of Kurdish People’s Protection Unit (YPG) forces, the al-Qaeda derived group formerly known as Jabhat al-Nusra which has been re-branded into the Front for the Conquest of the Levant (JFS), and the Islamic State. The list of aforementioned groups is nowhere near comprehensive and involves many shifting inter-group dynamics such as fighters who move from one group to the next, and clashes between formerly allied groups. Additionally, Turkish forces have entered the mix by launching a recent incursion into northern Syria to establish a safe zone in currently SDF-held territory. Foreign nations funding various fighting groups in the Syria war include the U.S., Turkey, some European countries, Russia, Iran, Israel, Jordan, Saudi Arabia, Qatar, some other Gulf states, and more.[3] The Syrian War operational picture is deeply multi-faceted, and looking at data on each of these groups through social network analysis would facilitate greater discussion and deeper understanding of the complex situation. Furthermore, visual displays of the data may help reach a wider audience with more easily digestible information.

Data, Methodology, and Network Measures:
àThe network I would explore includes all of the groups involved in the Syrian War. I would collect data from well-respected think tanks, NGOs working on the ground, and government factsheets on which groups operate where and have what capabilities. Additionally, I would include data from a social network analysis conducted of Syrian rebel alliances, published in the Journal of Conflict Resolution.[4] Network measures would include questions such as the following:
·       Group type (foreign nation forces, militant group, terrorist group, paid mercenary group, proxy group, etc.)
·       Overall allegiance? (Pro-Assad, Opposition, other)
·       Professional fighting force? (Yes/No)
·       Holds territory? (Yes/No)
·       Size of territory held (with different ranges to select from):
·       Territory held in which governorates (there are 14 total):
·       Air power capabilities? (Yes/No)
·       Funded by:
·       Provided materiel support by:
·       Overlapping fighters with:
·       Major ethnic composition:`
·       Major religious composition:
NOTE: The above categories would be fully defined, with the understanding that there may be some debate or overlap in categories.

What will the SNA help you do?
àThis SNA would help answer questions such as which groups have air capabilities and hold territory in Aleppo Governorate, and allow for visual displays of important network connections such as groups funded by both Iran and Russia.

[1] Hauer, Neil. “The Rise and Fall of a Russian Mercenary Army.” Foreign Policy (blog). October 6, 2019.
[3] Laub, Zachary. “Who’s Who in Syria’s Civil War.” Council on Foreign Relations. April 28, 2017.
[4] Gade, Emily Kalah, Michael Gabbay, Mohammed M. Hafez, and Zane Kelly. “Networks of Cooperation: Rebel Alliances in Fragmented Civil Wars.” Journal of Conflict Resolution 63, no. 9 (October 2019): 2071–97. doi:10.1177/0022002719826234.