“Predictive Modeling Is Here to Stay” Discussion on the LinkedIn Group: Workers Compensation Roundtable

Below is a response to my blog, Predictive Modeling Is Here to Stayin a discussion published on the Linked In group, Workers Compensation Roundtable last week.)

Responder 1: Predictive modeling is particularly useful on smaller policies. An account with millions of dollars in payroll and high premiums generally gets more attention so the quality of information is higher. Given the size, the history is much more predictive of future results. On the smaller accounts, you don’t typically have as high quality information and a lack of data on which to make sound assumptions. You have to underwrite the book more than the individual account. Predictive modeling allows a carrier to do that fairly, as you point out, by find those metrics that are real drivers or results.

Responder 2: Predictive Modeling clearly is being effectively engaged to improve the WC underwriting process. There is also a growing initiative to apply predictive models in the post claim world, using predicted loss costs from FROI data to create triage protocols, severity specific claim handling strategies, managed care referral triggers, and other resource allocation decisions. Finding the Pareto 20% quickly and deploying high case cost mitigation strategies can be achieved with predictive modeling tools. This is an equally important initiative for employers to engage with carriers and vendors managing their WC programs.

Me: Agreed! I will be covering how SI employers can use predictive modeling in a future post.

Responder 3: Annmarie, thanks for your posting. I followed through on your link, then to the other links…the article in Contingencies was hard to read online and I would very much like to see it in an easier to read format. It is my impression that predictive modeling has in effect been in workers comp for decades. What IT has done in the past 5 – 10 years, with more power, more accurate data, and far easier integration of disparate data sources, is to given real meaning to the word “model.” As the phrase goes, “all models are wrong; some are useful.” That is, one has to have the mind set to persistent in trying out different ways of looking into the future, always be willing (and able) to experiment. And this brings up for me the great barrier to predictive modeling in workers comp: the resistance of some claims and underwriting executives to challenge their own ways of predicting, which are crafted over many years and based on very flawed perceptions of reality. They thrive on anecdotes that they will not give up on.

Me: Part of predictive modeling is using new non-insurance-related data sets. This is definitely new. Barriers are being overcome. They must for insurers to stay competitive. One workers’ compensation insurer wrote me after my post asking how small-to-medium-sized carriers are handling predictive modeling when data is limited. More insurance-related data is becoming available. More workers’ compensation carriers are adopting predictive modeling and I will be posting something on this by the end of the week. Sorry about the Contingencies article. I like to refer people to the actual publication’s posting. If you go under my “work samples” tab on my blog, you will also find it there. To save time, I am giving you the link to it below. It is a paper copy I had scanned in. Here’s the link: https://annmariecommunicatesinsurance.files.wordpress.com/2012/01/workerscomppredmod-contingencies2.pdf

You are correct that computer technology is also making predictive modeling more possible. As I am covering in an article yet-to-be published, more carriers are installing new systems that support predictive modeling. Thanks for reading and for your comments! Stay warm!

Responder 4: (Responder 3) is right The “dinosaur” mentality based on years ( actually decades…) of “gut feeling” still prevails out there. It is in claims, sales, loss control, risk management, etc. Part of the problem is that people inherently resist “change”. 

When “the model” does not agree with the “historical perception” then “the model” must be wrong, because the person making that “right / wrong” decision has made those type of decisions based on a “gut feeling” (i.e. – experience) for the past 40 years….. 

When the data CONSISTENTLY swings in favor of “modeling”, then the “gut feeling” as the “primary” critera has to be “repositioned” as a secondary consideration, rather than the primary consideration. In time, as models improve, maybe 95% of WC claims can be successfully “modeled”. The remaining 5% are going to have to be reviewed by a person that determines that the model does not apply to this SPECIFIC claim and maybe a “gut feeling” decision is more appropriate, in this single case. At the end of the day, every claim is unique. 

WC modeling has come a long way in a short time, and like everything, will continue to improve as more valid data is compiled and analyized. You just have to convince “the boss” that the data is more accurate than 40 years of “gut feelings”……

Responder 3: Annmarie, thanks for posting the article in a readable format. It is the best discussion I have ever read on predictive modeling in workers comp. To be sure, you are light on the resistance to modeling such as (one responder) cogently points out. The first serious model I saw as introduced into a workers comp insurer about 2000 and was abandoned eventually because adjusters felt threatened. But your article is a gift to the community. My abiding interest in modeling took me to nutty extremes last September when I visited the grave site of the poet, Rilke, whose poems of 1900-1925 were, I feel, an aesthetic interpretation of modeling, in his case, how you change your whole life, time and again.

Me: Thanks for the compliment on my article.  Just want to address (Responder 3) and (Responder 4). For underwriting purposes, predictive modeling often quantifies what underwriters knew with their gut feeling and that alone is important. But it goes far beyond that. 

Workers’ compensation predictive modeling got started around 12 years ago by the largest wc insurers and applications keep growing. I do not agree that the “dinosaur” mentality is holding back insurers, or not at least like it once did. 

It can’t. There is just too much profit to be made, not to mention all the market segmenting opportunities insurers can use to fine-tune their competitive strategies.

The bandwagon is filling up and nobody wants to be left behind. 

Predictive modeling, as it is now understood, started with Progressive Insurance around 30 years ago. It forever changed how auto, and later other personal lines, would be underwritten by the entire industry. 

It will not take that long for the same to happen with workers’ compensation predictive modeling. Commercial lines actuaries and statisticians have learned a lot from the personal lines experience and there are more tools available. Automation and the latest computer systems support predictive modeling much better than those of the past. Some companies are developing computer systems with predictive modeling in mind. 

The article I posted does cover barriers to predictive modeling, but they are quickly being overcome thanks to good old fashioned profit motive. Workers’ compensation insurance will experience the Progressive Insurance phenomenon. If employers did not have a good enough financial argument for following all the great advice to reduce losses, they do now.

Responder 5: Excellent article. Re: your comment “predictive modeling often quantifies what underwriters knew with their gut feeling and that alone is important” – don’t underestimate the importance of this. Underwriters are trained in the Joe Friday, “just the facts” school and are hesitant to make a decision without something in the file. 

On the claims side, this is also true. I was involved in putting together a very simple model which showed UR denial rates or treating physicians – a simple provider scorecard. When the examiners saw this, they were delighted to have real evidence that Dr. X was a problem.

 Me: Thanks (Responder 5). I have been keeping my focus on predictive modeling as an underwriting/rating application because this is the most common and immediate way employers will notice a difference in premiums based on their experience. Claims predictive modeling is also picking up some speed and we can learn a lot from group health care in this regard. I have more articles coming out on predictive modeling and when they are published, I will share! 
Appreciate all the comments everyone!

(Note: Permission was granted by one of the group’s managers to publish this discussion with the names of commenters removed.)

Predictive Modeling Is Here to Stay

Credit: Brian Casey

Credit: Brian Casey

Last week’s blog covered how workers’ compensation predictive modeling will affect premiums. But there is much more employers need to know.

2) Predictive modeling is here to stay. Predictive modeling is well on its way to becoming an insurer best practice. According to a study released by Towers Watson last year, 72 percent of the comp carriers surveyed said they are either engaging in predictive modeling or soon will be. (For more info on the survey, please see my workers’ compensation predictive modeling article.

You can’t blame them really. Insurers engaging in predictive modeling are becoming more profitable more accurate rates, lower losses and better customer retention. Those not yet applying predictive models will have to follow suit to stay competitive. The latest Towers Watson survey is coming out soon and I expect it will show more insurers are pursuing predictive modeling in workers’ compensation. (I’ll let you know when it comes out.

… insurers engaged in predictive modeling are looking for that “secret sauce”…

3) Auditing parameters will change. In a perfect world, employers would be completely truthful when applying for insurance and understand the comp classification system to avoid misreporting. But the world is not perfect. The insurance industry estimates that 15 to 20 percent of employers are not paying their fair share for workers’ comp.

Traditionally, larger risks – those paying more than $10,000 in annual premium — got more attention from auditors. Insurers are using predictive modeling to locate employers more likely to need auditing. Using government sources, for example, insurers are looking at wages and employment information to reveal payroll information discrepancies.

Predictive modeling can also reduce the shell game played by shadier employers who change their company name to get a clean experience modifier or switch insurers to hide high-risk jobs under other classification codes.

4) There is a lot of experimenting going on.  Predictive modeling is relatively new to commercial lines. Insurers engaged in predictive modeling are looking for that “secret sauce” that will enable profitability.

This means there is a lot of experimenting going on. Insurers are experimenting by using new types of data as factor proxies.  They are also learning how much weight to give factors, the best combination of factors, and other considerations.

(Note: This is part II of my blog series on What Employers Should Know About Workers’ Compensation Predictive Modeling. To learn how predictive modeling will more accurately reflect an employer’s specific experience, click here.)

Next week’s blog : Predictive Modeling and Your Insurance Agent

Be the first to know! Follow this blog by clicking “follow” at the bottom right hand side of this page.

What Employers Should Know About Workers’ Compensation Predictive Modeling (Part I)

A special “thank you” to Mark Wells for making this blog a “manager’s choice” on LinkedIn’s Work Comp Analysis Group.

If predictive modeling is not already influencing your premiums, chances are it will soon.

Predictive modeling is a statistical approach that looks at traditional variables, such as employer’s history of losses, and adds non-traditional factors, such as the average salary of employees. Just as credit scoring is a proven risk indicator for personal automobile coverage, considering a company’s financial health can indicate the degree of workers’ compensation risk.

Factors, representing colors here, are fed into predictive models.

Factors, representing colors here, are fed into predictive models.

There’s a lot employers need to know about predictive modeling.

1) Premiums will more accurately reflect each employer’s specific experience. Predictive modeling is a more finely-tuned instrument than traditional pricing approaches. That’s because it gives weight to new and traditional factors.

This is great news for progressive employers doing everything they can to foster a safe and healthy workforce, boost overall morale and help injured workers return to work. Efforts such as more early claims reporting, which helps employees get better medical treatment and heal faster, are more quantifiable through predictive modeling and can positively affect premiums.

Conversely, employers who view comp as a price of doing business are in for unpleasant surprises. They could very well find themselves in the high risk pool.

Traditionally, workers’ compensation insurers priced on a “univariate” classification basis. This means insurers would look collectively at employers and their employees in the same class. An employer’s experience modification factor is then used to reflect its actual risk history.

Conversely, employers who view comp as a price of doing business
are in for unpleasant surprises.

Another factor can be the affordability and quality of health insurance coverage. This can determine the incentives for cost shifting medical treatment to workers’ compensation. Healthier employees tend to recover faster, which shortens the duration of workers’ compensation claims. Therefore, initiatives taken to improve employee health, such as wellness programs and disease management, which can help reduce co-morbidities such as diabetes, smoking and obesity, are also possible factors.

Other employer-specific factors that can be fed into predictive models including payment history, loss control initiatives, age of company, average age of workers and length of employment because these have been shown to affect outcomes. Income levels of employees can be an indicator of employee morale based on the assumption that better compensated workers are happier.

Personally, I am excited! Predictive modeling will give employers even more financial incentive to pay attention to employee health and safety on and off the job. That can only be a good thing.

This series is primarily based on articles I have published in two industry magazines. Check them out by clicking here and here.

 Be the first to know by clicking on the “follow” button on the bottom right corner.

Predictive Modeling ‘s Impact on Agents and Brokers—and Their Clients

Predictive modeling pieces together data in new ways

Predictive modeling pieces together data in new ways. (Thanks to steve.kargs.net for the pic.)

Leader’s Edge magazine just published my article on how predictive modeling will affect agents and brokers who sell commercial insurance – and their customers. The article, Modeling the Futureexplains how agents and brokers who invest in their own predictive models can see significant returns on their bottom line.

Many agents and brokers are already being affected by the predictive models of the insurance companies they represent. To a greater extent, predictive modeling reveals the true risk of employers. Some agents’ customers will see price improvements while others will experience the opposite. Insurers also need to do a better job informing their agents of how they are using the models so agents can respond appropriately.

Those who adapt quickly will do well, those who do not will be left behind.

Agents and brokers with the interest and resources to invest in their own predictive models will be able to offer better products and services. My article features one company that is increasing revenue by improving the value proposition of workers’ compensation products and services with predictive modeling.

Some agents bristle at the technology, but predictive modeling is here to stay. Those who adapt quickly will do well, those who do not will be left behind.

I hope you enjoy this last article under my current byline. Since I am getting married this Saturday, my future byline will be Annmarie Geddes Baribeau.

If you want to read more about predictive modeling, please check out the “predictive modeling” section under “topics.”

Understanding the Workers’ Comp Experience Mod Factor

Workers’ compensation premium costs are expected to continue rising. The state with the highest share of workers’ compensation premium – California  — proposed a 4.1 percent overall increase in comp insurance. (http://www.californiahealthline.org/articles/2012/4/13/workers-compensation-insurance-group-seeks-41-premium-increase.aspx

)The central reasons for the premium increases, regardless of jurisdiction, stem from the amount of claims (claim frequency), rising medical costs (which run higher than for general health care) and increases costs from claim severity.

Please note that claim “severity” should not be confused with the severity of injury or occupational illness. Claims severity, refers to claim duration and costs. Of course, injury severity can affect a claim’s length, but the effectiveness of claims management is also a major contributor.

Most claims are referred to as medical-only claims, which is generally defined as any claim where the injured worker losses less than seven days from work. In such cases, and depending on the employer, injured workers generally use their sick time.

Most employers do not know their workers’ compensation premiums are going up until they get their insurance bill. Higher premium can be a real shock to the chief financial officer trying to work with human resources to contain overall employee costs.

How can employers know ahead of time if premium increases are likely to come?

The actuaries are the first to know. Workers’ compensation is a long-tail line. That is, because it can take a lot of time from when claims are filed until they are financially resolved, there is a delayed effect to what actually happened in a given year. Actuaries are the first to see changes in cost and other trends that ultimately affect premium.

Most employers do not know their workers’ compensation premiums are going up until they get their insurance bill.

You should be very in tune with your company’s experience modification factor, which reflects the amount and severity of workers’ compensation claims for your business and its industry classification. Safety programs and other initiatives, which I will cover in a future blog, can go far in improving the experience modification factor.

Employers should also be aware that insurers are increasingly applying predictive modeling to determine more accurate premium. You are probably familiar with predictive modeling for personal auto insurance. Like credit scoring, it considers more than experience including education and occupation as indicative of driver safety and potential car accidents.

Predictive modeling in workers’ compensation considers new variables to determine loss potential. These include the overall financial health of employers, the number of employees being covered and more detail about the workplace location.

Using the correct combination of information about an employer, some actuaries believe, is a more effective way of predicting future losses than the experience modification factor. I expect insurers will change underwriting and pricing assumptions due to predictive modeling. This will also be covered in a future blog.

Predictive Modeling Integration: How Long Must We Wait?

Predictive modeling connects variables, making more predictive ones. (Drawing by Kristen Lipold, age 10)

Predictive modeling will change insurance forever! I got psyched about this when, for a current assignment, I re-read my article on workers’ compensation predictive modeling that was published about three weeks ago.

Truthfully, I rarely re-read anything I publish because I get tired of the topic. My goal for the article was write the most comprehensive story on workers’ compensation predictive modeling since most articles on the subject only cover individual aspects of it.

By reading it afresh, I could see all the pieces of predictive modeling coming together to create the ultimate  — predictive modeling integration! To explain, here’s a quick definition: Predictive modeling, also known as predictive analytics, uses data and analytical techniques to reveal and create the best predictive risk indicators. Credit scoring in personal auto, while controversial, is a good example.

In workers’ compensation, predictive modeling will transform best practices into employer necessities. Why? Because predictive modeling goes beyond influencing the experience modification factor. It is a more finely-tuned underwriting instrument, enabling insurers to better able to identify and reward the employers who are truly investing in workplace safety, workers’ compensation and return-to-work programs.

I believe integrated predictive modeling will help smooth out the famously cyclical workers’ compensation system. It will change everything from how injured workers are serviced to discussions at the public policy level.

We are not there — yet. I am not aware of any organizations that have connected different predictive modeling applications to maximize its potential. Barriers to implementing predictive modeling abound, but nobody has ever mentioned to me the most obvious: the lack of clear communication between professionals from different specialties!


I mean really, when was the last time you saw an actuary and a claims professional
having a beer and doing some serious integrative strategic thinking?

­­­In fact, because predictive modeling is not being communicated very effectively, everyone does not understand what predictive modeling is and what it can do. Insurance professionals are “siloed” in different departments and disciplines. Predictive modeling is the work of actuaries and statisticians who, as a profession, are not known for being clear communicators. 

Self-insured, self-administered companies could reach the ultimate in predictive modeling before insurers if they can get CFO and CEO buy-in. Such a commitment requires communicating a compelling business case, which is also a challenge.

However, insurers are increasingly seeing they must buy-in to compete and improve profit potential. So for this blog, I am focusing only on workers’ compensation insurers and their employer clients. Predictive modeling is most mature in pricing and underwriting. Premium auditing also has maturity because it looks at many of the same factors as underwriting.

Naturally, you would think professionals from the premium auditing and underwriting departments are hanging out at lunch and discussing how premium audit results should affect its future premium. But since these departments have traditionally been separate and apart, the missing dialog is denying this potential to too many insurers.

Claims predictive modeling gets me the most excited because it can detect adverse claim events, prompting attention that can ultimately reduce severity costs and help injured workers get better and back to work faster. The cost containment potential savings for employers is also fantastic! But claims people don’t hang out with underwriters, premium auditors and actuaries.

I mean really, when was the last time you saw an actuary and a claims professional having a beer and doing some serious integrative strategic thinking?

Since each group is struggling through its own predictive models without talking to one another, integrating predictive modeling has a way to go.

But once integration takes place, claims predictive modeling will ultimately provide different and better variables to create finer-tuned predictive models for underwriting and pricing. Then the business argument for employers to improve their workers’ compensation programs becomes even more compelling than the experience mod.

This is just the beginning of synergistic thrill of integrated predictive modeling. The excitement intensifies even more when predictive modeling is applied and integrated with benefits, such as health care, and other commercial insurance lines.

I feel so blessed to see the big picture! More to come in a future blog.

Workers’ Compensation Predictive Modeling

My article on workers’ compensation predictive modeling was just posted by Contingencies, the magazine of the American Academy of Actuaries (http://www.contingenciesonline.com/contingenciesonline/201205#pg37.)

It demonstrates how predictive modeling is changing everything from how policies are being written to how it can help injured workers receive the care they deserve after a work-related injury or occupational illness.

Personally, I am excited about the topic, which is why I suggested it. My special thanks go to Curtis Gary Dean, professor at Ball State University, Peter Wu of Deloitte Consulting, Brian Stoll of Towers Watson, Rong Yi of Milliman and Janine Johnson of Insurance Services Office Inc. Without their consistent availability, the article could not have been written!

Predictive modeling is the wave of the future in workers’ compensation! Ultimately, predictive modeling will affect our lives in various ways. Please check it out! Thanks! — Annmarie