A Fletcher Connected World: An Examination of how Fletcher Alumni Connect Businesses, Organizations and Institutions

Background

The strength of an alumni network is its connectivity. However, such networks are almost always examined as individual points grouped by characteristics. They tend to be depicted as pie charts sliced into sectors or percentages across different salary ranges. In 2011, Rubens et al. published a paper called “Alumni Network Analysis” that explains the need to preserve connectivity when studying alumni networks, and provides insights from social network analyses of Stanford, Harvard, MIT and UC Berkeley alumni. A paper titled “The Power of Alumni Networks – Success of Startup Companies Correlates with Online Social Network Structure of Its Founders” published in 2009 by MIT’s Sloan School of Management exemplifies the utility of an alumni analysis to answer more targeted questions as well. The researchers analyzed alumni networks from twelve German universities within Germany’s startups, measured which network had the most success, and examined the attributes associated with success. Both papers are from a few years ago but are still relevant because the way that alumni networks are typically visualized has not changed.

I’m interested in the infamous network of “Fletcher Mafia.” I want to examine the way that this network connects businesses, organizations and institutions accross the globe.

In the rare cases that examine alumni networks through SNA, they are built as two mode data sets that draw links between people and their workplaces (see figure 1, from Rubens et al. paper).

However, much more interesting than how alumni are connected through work is the access network that alumni create among all of their places of employment. The basic assumption in valuing an alumni network as an asset is that anyone who shares an alma mater will be willing to respond to thoughtful email from a current student or other alumnus. (The cynical way to support this idea is to think about the strength of a school’s brand, which means helping a fellow alumnus is actually self-serving. However, since any single alumni probably doesn’t have the bandwidth to prop up enough fellow alumni to impact the school’s reputation, it might be unfair to consider this as a very significant factor). Fletcher takes particular pride in the degree to which its alumni are willing to support each other. After only two short months here, I have already witnessed this on a number of occasions and find it hard to imagine there are very many alumni networks that go out of their way for strangers to the extent of the so-called Mafia.

Therefore, I would argue that every institution employing a Fletcher alumnus has a connection to every other institution with a Fletcher alumnus. From here on out, I will call the map of institutions connected by having Fletcher alumni an “access network.”

Following this chain of reasoning, it becomes necessary to go back and revise the first sentence of this blog post: “The strength of an alumni network is its potential for connectivity.”

Questions

Group 1: How connected is the world through the Fletcher alumni network? Stated differently, how broad/strong is the access network created by Fletcher alumni? Which sectors have broader/stronger access networks? Which global regions have broader/stronger access networks?

Group 2: How has this access network shifted (among sectors and globally) over time?

Group 3: How successfully is the Fletcher access network being leveraged? How successfully is the Fletcher access network being leveraged in different sectors and regions?

Group 4: To ensure the baseline assumption that the Fletcher network has particularly strong propensity to help each other (here defined as “access ties”), it would be worthwhile to compare the strength of the Fletcher alumni network to alumni networks of similar schools. I would probably start with the Kennedy School and Johns Hopkins’ School of Advanced International Studies.

Data

Career services has created a very detailed career portal (“Fletcher Career Central” or “FCC”) with the jobs, bios, contact information and year of graduation of hundreds of Fletcher alumni. There is also a Fletcher LinkedIn group. Additional information about specific people could most likely be found online, or perhaps by talking with career services in case the office has additional records not posted on the career portal. This data would certainly take a long time to collect and organize but would not be difficult to find. The greatest challenge will probably be ensuring that the present employer information is up to date.

I will have to make some decisions about how to record this data, such as whether to group big departments as one employer or to divide them into smaller divisions. For example, the FCC lists the US Department of State as a single employer with 43 Fletcher contacts, but also provides the option of clicking each of these contacts for their specific divisions. I will also group these organizations categorically. I plan to break them down by public sector versus private sector, by region, by type of organization (i.e., government department, research organization, think tank, nonprofit) and by focus (i.e., peacebuilding, development).

Methodology

This network will show every organization, business and institution where a Fletcher alumnus is currently working, consulting or receiving a degree. Each node will be connected to every other node. The strength of the access connection between any two places will be based on two factors. The first is the number of alumni at each place. Two companies should reflect a much higher access strength if they are each comprised of fifty percent Fletcher alumni than two that have just one alumnus each. The second is the separation in time that those alumni attended Fletcher. There are a few possible ways to categorize this. If hypothetical Fletcher alumna Jane graduated in 2002 and works at Root Capital, she has a very strong access connection to anyone who overlapped with her at Fletcher since it is very likely that she either took classes with them, went to events together, or at the very least has mutual friends. Therefore, I would assign a connection strength of 1 between Root Capital and organizations with alumni who graduated in 2001, 2002 or 2003. It is also likely that if Jane emails someone from the graduating class of 2004, that person will know people graduates from 2003 who knew Jane. I would label the bond strength between Root Capital and organizations with alumni from 2004 or from 2000 with a connection strength of two. One way to assign the rest of the bond strengths is to continue this trend and make the degree of each connection equal to the number of years between graduating classes. However, beyond the first few degrees of separation, the impact of an additional year difference in graduating class will probably get smaller. Therefore, it might make sense to groups these degrees in wider buckets, and say that anyone who graduated within 5 years of someone else has a connection of ‘3,’ anyone who graduated from 6 to 10 years of someone else shares a connection of ‘4,’ etc.

There are a few ways that this can be visualized. In addition to looking at connections in specific sectors or regions, these organizations could also be labeled with their latitude and longitude, allowing these organizations to be accurately spaced in visualizations and overlaid on a map.

Another consideration is time of employment. Although the principal network should show only current employees, it is often true that organizations tend to have sufficiently strong connections with recently departed employees. An expanded version of this network that includes employers who have had Fletcher alumni within the last three years might also be helpful. The group two questions, which ask about the change of the Fletcher access network over time, require the creation of a similar network with alumni attached to the jobs they held at whatever year is being examined. The group three questions necessitate a survey question asking alumni to identify all organizations they partner with that have Fletcher alumni. The access network would then be compared to the network of realized connections, with a particular focus on whether places with stronger access ties are more likely to have realized connections.

To answer the group four questions and test the strength of the Fletcher alumni against that of the Kennedy School and SAIS, it is necessary to devise a methodology for comparing alumni networks. Schools often present the advantages of their alumni network to prospective students by discussing how well graduates are positioned (e.g., 23 in the World Bank, 43 in the State Department, etc.). However, comparison based only on the success and placement of alumni only works under the assumption that the average alumni at any school is equally likely to respond to a thoughtful email from a current student or other graduate. This assumption is faulty. Any comparison between alumni networks should factor in a measure of the average “helpfulness” of that network. Alumni spread across 4000 different organizations are all useless if none of them are willing to take a phone call. To do this, all alumni should receive a survey asking which other alumni they have reached out to for professional help and then rate the helpfulness of each person on a 1-5 scale (1 signifying no reply, and 5 being on the level of a job offer). The densities of these networks could then be compared to see which have more frequent and more meaningful “help” connections.

Conclusion

This project would provide a visualization of how much of the world is connected by Fletcher alumni, and how this has shifted over time. It would also serve as a resource for alumni and their organizations to see where they can easily build or strengthen connections. This might also be seen as a helpful tool for the Fletcher alumni office and Fletcher admissions.

----

Innovation Ecosystem Network, Media X, Stanford University N. Rubens, M. G. Russell, R. Perez, J. Huhtamaki, K. Still, D. Kaplan, and T. Okamoto, “Alumni Network Analysis,” in Global Engineering Education Conference (EDUCON), 2011 IEEE, Amman, Jordan, 2011, pp. 606-611.

Nann, Stefan and Krauss, Jonas S. and Schober, Michael and Gloor, Peter A. and Fischbach, Kai and Führes, Hauke, The Power of Alumni Networks - Success of Startup Companies Correlates with Online Social Network Structure of its Founders (January 11, 2010). MIT Sloan Research Paper No. 4766-10. Available at SSRN: https://ssrn.com/abstract=1534699 or http://dx.doi.org/10.2139/ssrn.1534699