Today actuaries and insurers are able to apply predictive analytics in novel ways because of advanced technologies, larger data sets, and increased computing power. A recent Risk & Insurance article featuring Milliman’s Peggy Brinkman and Phil Borba explores four key areas where advances in predictive analytics are changing the way insurers conduct business: claims, driving safety, property risk, and competitive rating.
A new report by the AAA Foundation for Traffic Safety found that six out of 10 teen driver accidents resulted from being distracted. Here are some key findings from the report:
The most frequent potentially-distracting behaviors were conversing or otherwise interacting with passengers and cell phone use.
• Passengers were present in 36% of all crashes
o 84% of passengers were estimated to be ages 16-19; fewer than 5% were parents or other adults.
o Driver was conversing or otherwise interacting with passenger in 15% of crashes.• The driver was engaged in cell phone use in 12% of crashes
o Visibly using a cell phone in 8% of all crashes;
o Cell phone use appeared likely (driver looking at or manipulating something out of view of the camera) in an additional 4%.• Cell phone use varied significantly by crash type:
o Visible in 21% of road-departure crashes, not visible but likely in additional 13%
o Visible in 10% of rear-end crashes, not visible but likely in additional 8%
o Least prevalent in single-vehicle loss-of-control crashes (most of these involved adverse weather or surface conditions).• Drivers operating or looking at cell phones looked away from the forward roadway excessively – spent an average of 4.1 seconds out of final 6 seconds before the crash looking away.
• The driver exhibited no reaction at all before impact in over half of rear-end crashes involving cell phone use.
Distracted driving has key implications for the auto insurance industry. In the article “Distracted driving: Text-mining accident descriptions,” Milliman’s Phil Borba discusses how text mining can help insurers capture cell phone use at the time of an accident. The article also identifies major considerations regarding the use of cell phones on auto insurance premiums and claim adjusting.
The cellphone (and use of electronic equipment generally) introduces a new aspect into the operation of a vehicle and one which poses a challenge for setting insurance rates. Generally, cell phones are seen as distractions to the operation of a vehicle and are likely to increase the frequency and possibly the severity of accidents. Furthermore, the nature of the equipment may have differing effects on the frequency and severity of accidents. Steering-wheel and voice-activated controls for built-in cell phones may be safer than plug-in after-market equipment, while external devices (e.g., hands-free headsets) may be the least safe model.
Insurance premiums notwithstanding, more responsible use of cellphones may occur through state laws prohibiting or limiting their irresponsible use. Similar to the case with powerful vehicles that can increase speed beyond safe limits, state laws may impose some control over the irresponsible use of cellphones. Violations of state laws can carry fines, and the points on one’s license can increase the driver’s insurance premium.
Drivers’ use of electronic communication devices will also influence claim adjusting—in particular, assigning responsibility and liability when a distraction has occurred. However, the most commonly used data-capture reports do not enable the report-taker to report if the driver was distracted, or the nature of the distraction. As an alternative, claim adjuster notes, and other text reports, can provide a great deal of information on the circumstances attendant to an accident. These text-format data sources provide a great deal of information that can be tapped for deciphering the activities preceding, during, and after an accident.
Watch this video to learn more about how text mining can enhance claim analytics.
Predictive analytics enable organizations to identify their most profitable and expensive customer groups. These tools analyze business data and processes to help executives make informed decisions. The following videos highlight Milliman’s predictive analytics solutions.
Predictive analytics: Uncovering value in the data
• Milliman consultants Nancy Watkins, Matt Chamberlain, Peggy Brinkmann, and Sheri Scott consider how insurance companies can use predictive analytics to enhance their pricing and underwriting practices.
Hurricanes and analytics: A 21st-century approach to pricing
• Matt Chamberlain discusses how geographic information systems (GIS) can be used to price hurricane risk. To learn more about how geocoding can lead to more accurate pricing, read this article.
Improving claim analytics through text mining
• In this video, Phil Borba explains how text mining can reveal valuable information hidden in the narratives of auto insurance claims that could lead to improved underwriting practices.
Read these two articles to learn more about claims analytics:
– Predictive analytics, text mining, and drug-impaired driving in automobile accidents
– Distracted driving: Text-mining accident descriptions
Milliman Datalytics-Defense: A new approach to understanding defense costs
• Milliman Datalytics-Defense analyzes data related to litigation costs to help businesses develop more effective claims defense strategies. Milliman actuary Chad Karls offers perspective in this video.
For more information about Milliman’s predictive analytics solutions, click here.
Text mining provides a means of extracting new information from unstructured data. Milliman’s Phil Borba uses text mining to examine accident descriptions and can thereby identify instances of distracted driving—including cell phone use, radio use, and other distractions—that might not otherwise be visible. This is but one example of how text mining can extract hidden meaning that was not previously available to property and casualty insurers.
To read the video transcript, click here.
Phil Borba recently discussed his latest text-mining analysis that links drug-impaired driving to higher claim severity with Carrier Management. In this article, he explains how text mining can reveal valuable information hidden in the narratives of auto insurance claims adjusters that could lead to improved underwriting practices.
Here is an excerpt:
These tidbits of textual information about drug-impaired driving don’t just inform the claims-handling process, but they can also provide valuable input to underwriters, Borba says.
Finding that an individual has been under the influence of a drug or taking medications could provide a reason for a non-renewal, he says. “Or it can become a rating variable—almost similar to if you had been convicted of a DWI [driving while intoxicated]. You are then put under a different category,” Borba says, suggesting underwriting and pricing applications.
A report summarizing Milliman’s findings about DUID—driving under the influence of a drug—was published in the April edition of Milliman’s Insight publication. The report notes that other carrier benefits of identifying drug-impaired driving include better claims triage and assignment of liability.
“Finding that the other driver was DUID may be cause for a subrogation recovery against that driver, or provide enough additional evidence to increase the likelihood or size of recovery,” the report notes.
Explaining the claims triage benefit, Borba says that if carriers react quickly to any indication of drug use, then they will be able to reroute claims to adjusters who are familiar with different medications and know how to test for those. That’s important because the establishing that someone was DUID is a complex process, he says—much more involved than establishing a DWI.
Previously, Phil identified major considerations concerning the use of cell phones on claim adjusting and auto insurance premiums in his paper “Distracted driving: Text-mining accident descriptions.”
Text mining narrative data can help uncover valuable information for insurers that might not be captured in conventional data systems. In a new article, Phil Borba shows how text mining can identify auto policyholders who may have been driving under the influence of a drug at the time of an accident.
His analysis finds a measurable increase in severity when there was the presence of a medication, prescription, or illegal drug for one or more drivers in an accident. While the finding may not be new, the manner that Borba used to extract information from the text in the accident descriptions is novel. The mining of the text descriptions can help improve an insurer’s claims and underwriting practices in many ways.
This excerpt explains the methodology and results of the text mining analysis:
Narrative descriptions for the 7,000 NHTSA (National Highway Traffic Safety Administration) accidents were broken into phrases, and similar phrases were grouped together using analytical models. After removing prepositions and uninformative prepositional phrases, the result was a data file with more than 13 million phrases.
Next, we used four different themes for identifying the presence of a medication, prescription, drug, or illegal narcotic. First, we identified phrases with a “taking medications” theme. We joined phrases with the word “medications” that indicated a driver may have been taking medications. For example, we joined “on many” and “taking pain” to form “on many medications” and “taking pain medications,” respectively.
The second theme followed the same process, replacing “medications” with “prescriptions,” which gave us phrases such as “on many prescriptions” and “taking pain prescriptions.” These two themes produced approximately 1,100 phrases.
The third theme joined an action and a drug name. The result from these joins was a long list of phrases with “had taken [drug name],” “was on [drug name],” and so on, replacing [drug name] with the names of drugs. For the present analysis, we worked with 3,590 phrases with a drug name. The fourth theme was a list of 52 references to illegal narcotics that we considered red flags when seen on an accident description. This list included “cocaine,” “heroin,” and “marijuana.”
In sum, the first two themes were general references to medications and prescriptions, the third theme captured references to drug names, and the fourth theme was a list of illegal narcotics that we considered “red flags” for a driver being under the influence of a drug. For each theme, a binary (0/1) variable was created to capture whether the presence of medications, prescriptions, a drug name, or an illegal narcotic was mentioned in the accident description.
An injury was reported to have occurred in 73% of the 6,949 accidents in the NHTSA database. We found a reference to taking or being on a medication in approximately 16% of the accidents, and an injury occurred in 82% of these accidents. Similarly, we found a reference to taking or being on a prescription or a drug in approximately 6.5% of the accidents and an 80% injury occurrence for these subsets of accidents. Finally, we found reference to an illegal narcotic in 2.4% of the accidents and that an injury occurred in 89% of these accidents.
To read the entire paper, click here.