Luis F. Alvarez-Leon, Johannes Paßmann, Johan Söderberg

OBJECTIVE
Track the different types of financial transactions present in websites related to a particular issue.

METHOD
1. We chose the issue defined by the slashtag “/ows” in blekko, a slashtag-based search engine. The purpose of this is to find a culled list of websites that prominently features the “ows” term and are thus presumably highly involved with the Occupy Wall Street Movement.
2. The search results obtained in blekko were extracted and input into Link Ripper. This gave us a text file with links that we can then feed into the Harvester tool.
3. Through the Harvester tool we obtained the dataset of 133 URLs, which we exported into Excel.
4. In Excel we edited the URL list to delete duplicates and links that corresponded to the search engine. This gave us a total of 117 URLs left.

Secure Transactions
5. Now use the Source Code Search tool for the string https:// which indicates a secure financial transaction.

First Attempt
6. We got 75 URLs from this method. However 7 of these are marked “found 0 times”. Why only 7 and not the rest of the URL list? Why do they even show up in the search?

Second Attempt
7. We cleaned our URL list to delete doubles and broken links.
8. Ran that list in Source Code Search again.
9. We got 62 unique URLs with https:// which indicates a secure transaction. This might include other sites that don’t fit what we’re looking for like Twitter and Facebook.
10. List expanded to 199 https:// links in total. We assume This is because it includes internal links with mentions of https://

Database Cleanup
11. Now we remove Twitter from this 199. There were 54 Twitter links. 145 left.
12. Then we remove Facebook. There were 71 Facebook. 74 left.
13. Do same for Google/Youtube. There were 5 Youtube and 0 Google. There are 69 URLs left in the list.
14. List came up with some “doubles” that are actually replicates with a slash (/) at the end. We will cull the list manually.
15. We now work manually with the database organizing it by types of paying/transaction systems. The goal of this is to collect a sample of buttons (URLs, pieces of code, etc.) to get a sense of how many OWS are engaged in financial transactions. Additionally this will enable us to identify what are the primary means of fundraising that are used in the sample of OWS related websites.
16. By manually working through the URL list we assembled a database of OWS related websites and their associated financial transactions. The database is organized in the following categories:

URL: Name of the URL
ACTIVE: Is the URL active or inactive (binary variable 1/0)

TRANSACTION: Does it include a financial transaction? (binary variable 1/0)

PARTOFMOVM.: Is this URL part of the OWS movement? (binary variable 1/0). For those URLs with financial transactions we classified them according to the system they use. Each one is a category in the database and we included the URL for the donation website or, where this was not possible, the relevant piece of code of the donation button.

PAYPAL: Uses Paypal

WEPAY: Uses Wepay

NETWORK FOR GOOD: Uses Network for Good

LOCAL: This category indicates website doesn’t use an outside funding service.

ANALYSIS
17. Once we have assembled the full database we will perform different types of analysis to explore the connections between different OWS-related sites and payment systems. A first step is to find out how many of the total sample are making use of which systems. We did this by eliminating all remaining external websites and making a graph in Excel.

18. Simultaneously we wanted to analyze the correlation between traffic as a proxy for popularity and a website’s choice of payment system. For this purpose we assembled a traffic database from information obtained through SiteAnalytics (Compete.com). We merged the traffic results in Excel with our payment dataset. and analyzing it in Excel.

FINDINGS
1. We found that WePay is the single dominant payment system used by OWS-related websites.

2. Although Paypal is the second individual payment system within the group, it is largely overshadowed by the aggregate of other payment systems such as localized donations for each website and Network for Good.

3. Presumably the low use of PayPal can be attributed to OWS's dissatisfaction and ideological disagreement with PayPal's operations and their relationship with banks. This can also help explain the rise and intensive use of WePay as an alternative payment system.

4. In the traffic analysis of OWS websites there was one clear leader (www.occupywallst.org) with more than a million visits in Oct 2011. The runner-up, (www.occupytogether.org) was significantly lower with just over 300,000 visits in the same month. There were three other sites that stood out from the rest of the sample: http://wearethe99percent.tumblr.com/ (294,548 visits), http://nycga.cc/(157,579 visits) and http://15october.net/ (85,736 visits). The rest were mostly sites referring to local OWS branches for several cities in North America and one in Tokyo, which is now inactive and redirects to a different OWS site.

5. Out of the top 5 sites by traffic, it is notable that only www.occupywallst.org (the top site) and www.nycga.cc (the site for the New York City General Assembly, in 4th place) have any donation buttons in their source code. The rest are not visibly involved in any financial transactions. From this we infer that the top sites in terms of traffic within the OWS movement are for information and diffusion purposes mainly. The collection of funds seems to be redirected and concentrated within the various local branches of the movement. This is consistent with the decentralized nature we have observed in the formation of OWS and its different offspring.