Let’s start with a line from Theodore Roosevelt: “Nobody cares how much you know until they know how much you care.” Now let’s ask ourselves—does this have any application to comp claim adjusting? According to a recent article in workerscompensation.com, the answer is a resounding Yes.
Obviously, we are all concerned about how much our adjusters know. After all, comp is a wonderland of intricate laws, regulations, procedures, and forms, often overlaid with new ideas during past years and varying in challenging ways from state to state. But what about the caring part? The author, Dr. Claire Muselman, opens her essay with a challenge:
Empathy is a concept so needed in the workers' compensation space and yet we have not invested the time, talent, and resources to further examine and develop this concept. Empathy: a core fundamental to build connection and trust.
She goes on to note that the social critic and philosopher Jon Gordon once said, “The story you tell yourself depicts the life you live and the actions you will take.” In other words, if we bring distrust and cynicism to a new comp claim, our experience in administering that claim will appear to validate our attitude, possibly regardless of the actions of the claimant.
Now is an especially important time to be exploring the role of empathy in claims handling. Across the insurance industry, the long-time “crusty claim professionals” are retiring in droves and being replaced by new adjusters, just learning the ropes. If empathy for the injured person can improve the claim process, now is the time to inculcate that idea before attitudes become hardened. Every claims organization has a culture. Dr. Muselman urges us to be brave enough to ask questions about the impact of the injury on the person’s life and to have the courage to listen to the answers. Connect the injured person’s emotions to the work injury. What transpired? What relationships have been stressed, strained, or broken as a result of this injury?
Perhaps the most important qualification for understanding the comp claim process at a human level is to have had a life-changing work injury of one’s own, to have been on the “other end” of the process*. Failing that, how can we introduce a proper quality of empathy into claims handling, letting the injured person know that they are neither the enemy nor merely an object to be manipulated?
Teddy may have been right about caring. Just maybe it is important. Speaking of “teddy”, at GB we have our own Gentle Bear**. We got rid of the harsh sounding old term “claim adjuster” and replaced it with “Resolution Manager” which gives a much better idea of what the job really entails. Adjusting is the province of chiropractors; the job at GB is finding the right resolution. That’s the difference between talk and a real culture of empathy and cooperation.
Finally, think about this:
We have to teach empathy as we do literacy. It is a learned skill. If everyone had it at birth, we would live in a different world.
*Been there. Your faithful correspondent was scheduled to be permanent and total back in 1989 after a work-related airline accident.
**Don’t know about Gentle Bear? Go to GB Gentle Bear : Gallagher Bassett for the whole story, going all the way back to his origins in Australia.
So What Was Your AI Thinking, Anyway
True: back in the early 80s, a personal friend who was also a major modern American poet* embarked on an experiment to teach his then state of the art computer to write lyric poems on its own. Among other works, it wrote this very short poem: I become anxious/ when I think about/ my electricity. Computers are like children—they learn what we teach them.
That brings up a few issues. Last fall the US Food and Drug Administration (FDA) issued the “Artificial Intelligence/Machine Learning (AI/ML) Based Software as a Medical Device (SaMD) Action Plan” from the Center for Devices and Radiological Health’s Digital Health Center of Excellence.** The FDA Action Plan itself is available here. We call your attention to this development because it seems highly unlikely that this action plan for SaMDs is likely to be the last such plan issued by one part of the government or another addressing the development and use of AI/ML applications. Only organizations actively involved in providing medical/health services appear to be addressed by this particular document, but the Action Plan follows on a number of studies which have shown that the use of AI/ML can carry some important hidden risks.
On its face, the AI/ML process would appear to be an effective and essentially benign process when used to suss out often hidden relationships between various factors in a data assemblage in order to understand deep down rules and probabilities of interactions. One small example from our personal experience in developing a claim complexity prediction system is this: the injured employee’s own subjective assessment of their comparative state of general health is very strongly predictive of whether the comp claim will resolve normally or become an outlier.
But—AI/ML operates on top of existing data sets often based on processes and procedures which have been in place for years. What statistical kinks may be hidden inside that data set? It wasn’t too long ago that the FDA itself learned that tests of new drugs based on results for groups of patients who were largely or entirely male don’t always tell the whole story for female patients. The ever recurring question is—do certain underrepresented subsets of patient (or claimant) populations have atypical results?
While the FDA Action Plan focuses on the operating systems for medical devices, the same question hovers over any and all deployments of AI/ML, such as uses for resume evaluation or employee management and retention. If you are guessing that the deployment of any AI/ML system based on a skewed data resource may constitute a potentially huge risk, you must have been looking ahead. Everything starts somewhere and a lot of rethinking about building models using AI/ML processes is getting kicked into high gear by the FDA and its highly visible Action Plan.
If your organization has already deployed or is seriously planning to use new AI/ML systems for important corporate functions, have you thoroughly vetted the data resources being used to derive the models and relationships between metrics to eliminate biases or results skewed by over/under representation of certain groups***? Has your internal team or your vendor team rigorously tested and cleaned your big data trove before deriving relationships from it? If not, some unpleasant results could be lurking inside those very sophisticated, almost magical mathematical models. We generally think of math as transparent and unbiased—and it is. But what about the data derived from old manual/semi-manual processes which may have been anything but transparent and unbiased?
We have a presentation we give on why it’s so hard to predict the future. Our page on using big data is very relevant here:
*His work is now standard in anthologies of 20th Century American poetry.
**A mouthful only a bureaucrat could love.
***Or results based on now obsolete practices such as redlining?
Quick Take 1:
Keep On Trucking, Ma'am
Back in the 70’s, your faithful scribe drove a GMC cab-over diesel tractor with a 14 speed Spicer transmission. Driving it was unforgiving and required long arms, strong legs, exquisite timing (manual transmission), and a head on a swivel to check the door sized mirrors constantly. And that was before I had to back it up. Big rigs today are much more user friendly, which is a good thing. The great resignation plus the national wave of boomers retiring is leaving, by one count, some 80,000 trucks without drivers.
One possible solution?
With women making up only 7.8% of all truckers, gender inclusivity is more than good corporate policy in the industry — it’s mission critical. To that end, technological innovation is driving change to make trucking more appealing for female job seekers.
Do you have a fleet in your risk management assignment? If so, you might want to read through the article we quoted above. Bringing in female drivers to keep the trucks rolling involves more than agreeing to interview the women who apply for driving jobs. For example, are the ergonomics of your trucks suitable for female drivers who tend, on average, to be smaller than males?
If you have long haul routes, what about life on the road and potential hazards? Do you subscribe to the Women in Trucking Association’s app Engage, a community platform for members to share advice and ask questions? Do you use the other resources from this organization, such as educational videos in which women truckers demonstrate various safety and loading procedures in ways meaningful to women?
For more reasons than we can list here, life after the pandemic requires a great deal of new thinking, of using new resources, of asking new questions. Not too many years ago, the phrase “truck driver” conjured up a very particular image. Today we file that image under “Paleolithic”. With 2022 technology, anyone who can drive a car competently can be a truck driver with the right training and experience. The good news is that keeping your trucks on the road just got easier.
Trucking—it’s not a boy’s club anymore.
Quick Take 2:
How Much Is 71 Centimeters
Not enough when it’s the depth of the Rhine at Kaub in Germany (28 inches in American units), a natural bottleneck for Rhine barge/ship traffic. Here in the US we’ve been watching the shrinking Colorado River which supplies both water and hydropower to a large swath of the American Southwest, but the Rhine is different. It is the superhighway of commerce in Germany. Everything from raw materials and coal to semi-finished and finished goods are shipped on the Rhine.
Or at least they were. The Rhine is drying up, a victim of the outrageous heat and lack of rain enveloping Europe this summer. As it happens, 71 centimeters stops most shipping. From a risk point of view, this has huge consequences. The last time the Rhine got almost this low (78 centimeters) it was credited with inducing a mini-recession in Germany. Deliveries of critical supplies and parts were late and the delays got pushed right down the line. If your company has operations in Germany or depends on materials from Germany, you may soon be learning even more about the Rhine.
Meanwhile, barge transport in the US is valued at over $40B annually. What is happening in Germany can happen here. Barge transport on the Mississippi was squeezed a few years back by low water, a problem which goes back to Mark Twain* and before the Civil War. While barge transport is efficient—one standard grain barge carries as much as 58 tractor-trailer rigs—it is weather dependent, as the situation in Germany shows**. Risk? What risk? Well, there’s an old saying in German we all might want to keep in mind. Wer baut auf Wind, baut auf Satan’s erbarmen: he who trusts the wind (weather generally) trusts the mercy of the Devil.
*”Mark twain” means two fathoms—a safe depth of water for a steamboat on the Mississippi, just like 78 centimeters on the Rhine. Sam Clemens appropriated this phrase as his nom de plume (pen name).
**For a broader view of the gathering weather risk here and abroad see Paul Carroll’s item at ITL: Is the 'Heat Apocalypse' an Insurance Apocalypse? | Insurance Thought Leadership
Say It Isn't So...
Ever had a tele-bot put you on hold? Ever wonder whether that big company you need to talk to has any human employees left—anywhere? Ever wonder whether a customer service webpage was designed on another planet by beings who have read about humans but never actually met one? Welcome to the 21st Century.
If you answered “yes” to any of these questions, you will understand why I was so delighted to see the topic of ITL’s on demand webinar: How To Stop Annoying Your Customers. I’m a long-time admirer of ITL’s Paul Carroll and his uncanny ability to put his finger on the topic of the moment.
Annoying your customers is an elemental part of risk, isn’t it? While this webinar focuses on the insurance industry, its basic lessons have a much wider application. I recommend the webinar as a 45-minute refresher on how to work with your clients. After all, risk management’s clients are the rest of your organization. Might be some useful ideas in here.
Words to Remember:
See that bright, blue-white star high in the heavens in tonight’s midnight sky? That’s Sirius, the Dog Star. It’s really two stars very close together some 8.6 light years from Earth and it’s part of the constellation Canis Major (the Big Dog). It dominates the night sky in mid-summer, hence the term—the dog days of summer.
Not that long ago our ancestors told time and navigated both the seas and the deserts using stars like Sirius. The light pollution in most cities is now so bad, we almost forget they are still up there. If you’re out late on a clear night in the next couple of weeks, look for Sirius. It is an old friend to mankind, our companion since we first learned to look up and see the heavens many eons ago.
Hear how Homer uses it to describe Achilles on the windy plains of Troy:
Priam saw him first, with his old man’s eyes
A single point of light on Troy’s dusty plain.
Sirius rises late in the dark, liquid sky
On summer nights, star of stars.
Orion’s Dog they call it brightest
Of all, but an evil portent, bringing heat
And fevers to suffering humanity.
Achilles’ bronze gleamed like this as he ran.
— Homer, The Iliad, c. 600 BCE
The summer sky map