January 17, 2020
Pindrop for Amazon Connect | A Balancing Act
Defending the phone channel presents various challenges, especially due to…
Are you at risk? Use our interactive tool to view the level of activity from fraudsters in your county. Either enter your zip or county name or simply click on the map. [iframe src=”/wp-content/themes/zap-installable/visualizations/irs-scam-counties/index.html?v=4″ width=”647″ height=”640″]
In our previous blog post, we recorded our interaction with the scammers by posing as the victim. We wanted to understand if the scammers are targeting individuals at random or are they targeting a particular demographic. We used our Phone Reputation Service databases to determine the location associated with the complaints associated with this scam.
We plot the number of complaints by county on a map of the US; the darker the color of the county, the higher the number of complaints. Right away, some locations jump out at us. Several counties in Florida look quite dark, as well as the NYC, DC and some other big cities. Our initial assumption was that the number of complaints would match up well with the population of a region. We do see that to some extent – NYC, LA, Miami and other big cities show up prominently. However, several smaller counties in Florida (Lee, Collier) and California (Marin, Merced) also stand out. What’s different about them? The call that we recorded with the scammer gives us a big hint. At one point the scammer specifically asks whether the intended victim was “born and brought up” in the US. They clearly indicated that the scammer preferred targeting people who were foreign-born. We were curious whether this was a pattern that explained the distribution of targeted counties.
The American Community Survey (ACS) collects extensive data about the place of birth of people in the US to estimate the foreign-born population in each. As shown in the histogram, the general distribution of foreign-born people in US counties is very skewed. Most counties have a very small fraction of immigrants, in fact the median is about 2.5%, and only 5% of counties have a foreign-born population of 15% or more. The table below shows the top 25 most targeted counties in decreasing order of number of complaints, along with estimates for the overall and foreign-born population. Right away, we notice that the all of these counties have a very high proportion of foreign-born people: from 3X of the median (Baltimore, MD) to 20X (Miami-Dade, FL). Additionally, many high-population counties (NY, LA, DC, Harris) are prominent on this list.
|County||State||#Complaints||Total Pop||Foreign-born Pop||%Foreign-born|
|New York County||NY||3087||1619090||455248||28.51|
|Los Angeles County||CA||2094||9962789||3473930||35.30|
|District of Columbia||DC||1598||632323||82048||13.54|
|San Diego County||CA||365||3177063||720485||23.24|
|Santa Clara County||CA||349||1837504||658753||36.83|
What is the relative impact of each of these two factors? In order to tease that out, we plot the total number of complaints on a bubble scatter plot, with the two factors, immigrant proportion and overall population as the axes. The size of the bubble indicates the number of complaints from the corresponding county. As shown in the chart, as the bubbles get larger, they are clearly closer to the Y axis (proportion of immigrants) than the X axis (overall population), implying that the presence of a high concentration of immigrants has a high correlation with the top counties targeted in this scam. This analysis clearly shows that the scammers are following a pattern of targeting immigrant-dense areas of the country. We started this investigation with our employee having a conversation with a single IRS scammer, and eventually demonstrated a consistent pattern of scammers targeting immigrants on a large scale. The main takeaway: if you live in an area with a large immigrant population, be careful when you get calls from unknown numbers.
Raj Bandyopadhyay, Peter Casanova, Philip Thrasher