Storytelling with Data Visualization

In today's world of "Big Data" everyone is looking for interesting patterns in their data and visualizing the results.  They think "pretty pictures" are the goal.  But, pretty pictures are often pretty useless if they don't reveal a story.  In order to understand what the patterns are emerging we need to understand the story behind, beneath and within the data.  The standard Who/What/When/Where/How is a good place to start.  Below is a story of how a client of Orgnet, LLC applied financial and social network analysis to tell a story of corruption and fraud to their city prosecutor.  The NGO gathered meta-data about financial flows and business relationships involving local slumlords.

* * *

Our client is an NGO (non-profit) that focuses on housing issues in a large US city.  They had been working with local tenants and soon started to notice some troubling patterns across a group of rental properties.

• raw sewage leaks
• multiple tenant children with high lead levels
• eviction of complaining tenants
• utility liens on the buildings

The NGO began to collect public data on the properties with the most violations. As the collected data grew in size, the NGO examined various ways they could visualize the data making it clear and understandable to all concerned. They tried various mind-mapping and organization-charting software but to no avail -- the complex ties they were discovering just made the diagrams hopelessly unreadable. They turned to social network analysis [SNA] to make sense of the complex interconnectivity.

The data presented below is not the actual data from the criminal case. However, it does accurately reflect the social network analysis they performed. The names and genders of the individuals, as well as the names of real estate holdings [LLC] and other businesses have all been masked. This case will be presented in the sequence the NGO followed, first they looked at the real estate holdings, then the owners of the holdings, and then their connections, which led to other connections, and more people and entities.

The NGO worked with the tenants and city inspectors to assess the buildings and document the violations. But every time documented problems were delivered to the current LLC owners by city officials, nothing would happen. When the city's deadline approached to fix the violations, the old LLC owner would explain that the property had changed hands and they were no longer involved. The buildings continued to deteriorate as owner after owner avoided adressing the violations.

Figure 1 below shows how a building came under new ownership. The gray links show the "sold to" flow as building ownership changed from left to right. Every time a property changed hands, it became a new LLC [Limited Liability Corporation] with new owners.

Figure 1

The blue links in Figure 2 show ownership/business ties for each LLC. This data was gathered by the NGO from public records. Everything appears normal -- a different set of players in each LLC.

Figure 2

Yet, things were not normal. The NGO discovered that some of the LLC owners were married. As the NGO peeled the onion, more family ties were found within, and between, the LLCs.

Figure 3 shows us that these LLCs were not as separate as they first appeared. The dark red links reveal family ties found in public records. The LLCs were not independent business entities. The business transactions were happening within extended families! A conspiracy was coming into focus.

Figure 3

The dark red links in Figure 3 reveal two family clusters. Yet, there was a curious gap -- the transaction between ghi LLC and jkl LLC. Where these clusters connected? How? These questions soon led to a key discovery: the mastermind behind the conspiracy. Conspiracies often work in this way -- masterminds are 2 steps, or more, from the events they planned. See our blog post on how network distance is used to hide actual intentions -- indirect quid pro quo.

Figure 4 reveals the family matriarch and patriarch. The matriarch (Heather) was discovered in public records, explaining the gap. Then her current husband (Moe) was a quick deduction. The gap turned out be the dividing line between Heather's first family and her current family. She was the point of overlap between the two groups. 

Figure 4

Once Moe was uncloaked, the NGO's chief investigator decided to explore how he was connected -- what other business ties did he have? It turned out that Moe had ownership interests in several restaurants throughout the metropolitan area... and he was on the board of a mortgage company.

Figure 5

A mortgage company? It was not just any mortgage company, Moe was on the board of the mortgage company that had financed many of the real estate transactions we have been following here. Moe's ties completed the connections of the conspiracy -- the "circle of deceit."

Figure 6 shows the complete conspiracy. It was now obvious that properties exchanged hands not as independent and valid real estate investments but as a conspiracy to avoid fixing the building violations. The green links represent borrowed money flowing into the buildings through new mortgages. As time went on, and the buildings appreciated in value during a real estate boom -- loans from the mortgage company allowed the owners to "strip mine" the equity from the buildings. This is a common slumlord modus operandi -- they suck money out of a building rather than put money back in for maintenance.

Figure 6

Network analysis is not just about maps. Once a map is drawn, you can measure it. Social network metrics reveal much about the nodes, and the clusters they form. Who knows what is going on? Who wields power or influence? Who is a key connector? Who is in the "thick of things" in this conspiracy? Our metrics reveal Moe and Heather are most integrated nodes in the network. The hi-lited metrics in the Report window in Figure 7, showing the InFlow software, provides mathematical support to what is quite obvious in the diagrams. InFlow allows us to quickly see the relationships between "the maps and the metrics" -- the pictures and the numbers.

Figure 7

The city attorney combined the network analysis, along with the city's own extensive investigation and was able to get a conviction of key family members. Later, all of one building's tenants filed a civil suit using much of the same evidence and won a sufficient award to allow all of them to move out into decent housing. Several tenants used a part of their award to start businesses.

The common wisdom is that only big business and government use social network analysis. Yet, there are many individuals and groups that are learning the craft, and solving local problems. Although social network analysis can not be learned by reading a book, it does not require a PhD either. Any intelligent person, under the right guidance, and with the proper tools, can apply the methodology to an appropriate problem and gain enormous insight into what was previously hidden.

What story does your data tell?