A recently-released analysis of workers’ comp claim frequency tells us what we’ve known for years – the percentage of workers that gets hurt on the job has been and continues to drop.

Yet one major insurer indicates there are warning signs that frequency may be ticking back up, albeit in a tightly defined sector of the economy. More on that below.

There are many theories about why frequency has declined for decades – more automation, more emphasis on safety programs and loss control, less heavy industry here in the US, low investment in infrastructure leading to fewer jobs doing heavy construction. Many theories, but I have yet to see any credible research into exactly why frequency is declining.

This is one of those data points that is enormously important, yet it doesn’t get enough attention. So, here’s the skinny.

Claim frequency is a percentage  – the number of injuries compared to premium dollars or FTE workers. Therefore the number of work comp claims is driven by the denominator; if premium or employment goes up, that can offset a decline in the percentage of claims.

But employment is about maxed out, so any changes in the percentage of claims will closely mirror the actual number of claims.

Another way to track the number of claims is to compare it to Federal data on the actual number of occupational injuries and illnesses. The graph below shows that the actual number of claims per 100 workers…

You’ve already figured out that the graph ends in 2013…so what about the intervening years?  Fortunately NCCI provides ongoing research into just that here.  When you dig deeper, we learn that total frequency dropped almost one-fifth from 2011 to 2016, led by office and clerical job classes.

As we learned at AIS in May, NCCI estimated frequency declined another 6 percent in 2017; the average decline over the last two decades has been 3.7 percent.

But.

Last week the Hartford announced it is seeing early indications of an uptick in claims volume. Chairman and CEO Chris Swift said [emphasis added]:

“workers’ compensation 2018 frequency trends are elevated from expectation…

our frequency in small commercial and middle market has turned positive this year. Based on our business and economic analysis, we view this trend as broader than just our book of business.…

Many businesses are struggling to find qualified employees and beginning to add more new workers to their payroll, generally increasing the risk of workplace injury versus what it would have been say a year or two ago.

Additionally, the tightening labor market produces more hours work for employee often resulting in fatigue and less training time compounding the risk of injury for the less experience workers.

Our uptick in frequency change has been moderate turning positive on a rolling 12-month basis. The actual frequency levels are now comparable to what we experienced in 2016 which is a very manageable shift in a book of business as large as ours.

The frequency increase is more pronounced among less tenured employees and it can be several times that of experienced workers.”

Notably, the Hartford attributed a 3.5% increase in Accident year combined ratio for its middle market business to this increase in frequency.

Couple of key points about this.

1.  The Hartford is the largest national seller of work com policies to small employers, and thus has the broadest lens.
2. As a major writer, it also has lots of dollars to invest in business analytics – so it knows more faster than many insurers do. This from Hartford President Dough Elliott:

we have now installed a new claim platform over our 5,000 desk throughout claim. And the ability to access what I’ll call structured data and to slice and dice and be on top of it and to look at your metrics and watch your trends is much advanced from where we were five years ago. And so we have monthly and weekly discussions but we’re sitting on top of trends that candidly five years ago were very manual in nature to try to get our arms around and they were slower than we’d like to them to be.

What does this mean for you?

• This is not unexpected; we are close to max employment; small employers are desperate for workers and don’t have the time/expertise/resources to screen/train/protect those workers.
• Time to look at your data – closely.
• And likely time to dust off those underwriting, safety, and loss prevention manuals.

 

Over the last decade I’ve worked with over 30 investment firms on perhaps 60 deals.  One question I almost always get is:

Would you buy this company?

And a related question:

What would you pay for this company?

For years I tried to answer those questions, factoring in the company’s service reputation; the uniqueness of it’s services and/or business model; experience of management; value delivered to it’s customers; and a bunch of other stuff.

I finally realized those criteria often had little to do with the “value” defined by most investment firms. And much more to do with Dutch Tulips.

To most private equity firms, “value” is what can they sell the company for in a few years. One would think the selling price would be driven in large part by those other criteria; in many cases, one would be wrong.

Recent valuations of some work comp service companies are – in my view – completely disconnected from the actual value inherent in these companies – actual value defined as the value they bring to their customers and the potential for those companies to grow and prosper.

In fact, what seems more important than actual value is the ability of the seller to craft a story about how the company is going to grow, it’s unique business model, it’s scalability and potential to be a platform to which other acquisitions can be added. This is typically future-forecasting, theoretical stuff based on assumptions thin enough to blow away in the slightest of headwinds.

But more often than not, enough potential investors buy into the story to create a bit of a feeding frenzy, until one agrees to pay way more than that company’s actual value.

I’m far from an investment expert – one look at my personal portfolio will prove that – and in no way am I saying the brilliant folks at private equity firms don’t know what they are doing.  Far from it – these people are doing exactly what they are supposed to do – make gobs of money for their investors.

What I am saying is these firms are rewarded when they sell the companies they bought for a lot more than they paid. That works out really well – if they can find someone to buy it at a hefty markup. At some point the next owner – or the one after that – finds out that the actual value of that company is far less than they thought.

It’s also known as the Greater Fool Theory, or the Dutch Tulip problem. You know the asset isn’t worth what you’re paying, but you’re sure you can find someone else who will pay more than you did.

What does this mean for you?

Beware of tulips. They flourish until winter comes. And winter ALWAYS comes.

Several years ago it was smaller private equity firms who bought companies for $75 – $200 million, built them up, and sold them off. Then it was the bigger PE firms who put together a couple of these and hoped to sell them for more than they paid – and were mostly successful.

One of the first deals was way back ten years ago – and what’s transpired with that deal is emblematic of the evolution of the industry.

PMSI was sold by owner Amerisource Bergen after the PBM industry’s founding company just kind of melted away. HIG Capital picked it up for  about $50 million, and quickly brought in one of the most talented managers I’ve ever had the honor of knowing – Eileen Auen. Greatly aided by a major cash infusion, Eileen and her team turned the all-but-buried company around. HIG profited  handsomely when it was sold to Stone River and merged with Progressive Medical, a deal that was backed by Kelso.

The new company, renamed Helios, grew and was eventually sold to OptumRx, a division of giant health insurer UnitedHealthcare.

Optum is what is known as a “strategic” buyer. Strategics are investors that buy assets to add to its existing operation, ideally leveraging existing infrastructure, customers, systems, and suppliers. The result is infrastructure costs (and ideally supply costs) are reduced and revenues increased.

So, we have strategics such as Optum, Mitchell (see acquisitions of Pharmacy Benefit Managers), and Paradigm buying companies that somehow make the whole greater than the sum of the parts.

I’d note that some of these transactions seem a bit of a stretch, but, hey, I’m not the one who went to Harvard Business School, so what do I know.

The financial investors are the other entities still in the game.  Of late, it’s been the giants – Carlyle, Stone Point, KKR – investment firms with billions in ready cash, deep knowledge of the business, and the savvy needed to pick out management teams and business models that will flourish in what is a declining industry – workers’ comp.

We haven’t seen much activity on the small end of the deal world – except for Adva-net and a couple smaller transactions things have been pretty dull of late.  While there are a bunch of smaller companies in various geographic and/or service niches, there’s been precious little trading activity over the last couple of years (with a couple exceptions)

There’s one more investor type that’s getting more active these days – distressed asset investors. Their strategy is to find companies that aren’t doing well financially and buy them – sometimes out of bankruptcy or just before they tank.

I’d expect we’ll see more of the distressed asset folks making the rounds in Las Vegas in six weeks, finding out what’s up and who’s making it and who isn’t.

What does this mean for you?

The work comp service industry is getting awfully mature.

 

My post last week about the (minor) impact of rideshare services on work comp elicited responses from a couple of transportation companies, as well as one anonymous reader who wanted to talk about OneCall’s program (more on that below).

Full disclosure – a few years ago I was among those asserting rideshare companies would disrupt work comp transportation. I’m a lot smarter now (well, perhaps not smarter, just able to see why I was wrong)

Jim Nygren of HOMELINK sent in a detailed comment, which led to a rather extensive back-and-forth email conversation (I disagreed with some of his assertions, and wanted to have a conversation instead of putting up the comment and responding). As readers know, I’m happy to get comments and corrections, but not enthusiastic about marketing masquerading as comments or contributions.

The Jim and Joe conversation was professional, courteous, and perhaps a bit frustrating for both parties. After wrestling with the specifics of our conversation, I decided to post it here.  I don’t have any business dealings with HOMELINK; I’ve heard mostly good things about the company, I do like that it is employee owned. I also appreciate the work that HOMELINK has done in an effort to use rideshare services; they’ve clearly spent a lot of time and resource on this.

Here’s the conversation (“Jim” denotes his statements, rest should be self-explanatory)

Jim – Appreciate your post and perspective as always.  There are some materially inaccurate comments/assumptions made that need to be addressed, my comments below.  The most important thing to know is Rideshare is NOT an exclusive option, if properly implemented it is a tool within a complete transportation solution.

Joe response to Jim comment – I never said or implied rideshare would be an “exclusive option”.

From the original post –

Here’s why I don’t think they’ll replace work comp-specific transportation companies.

This is not to say L/U (Lyft/Uber) won’t be part of the answer – I’m just suggesting that “part” may be pretty defined and not nearly as big as we may have thought a while back.

Jim – My comments come from the perspective of the company I work for, HOMELINK, which began offering Rideshare as a solution in 2015.  I look forward to your feedback.

(Jim then quoted text from my original post, with his comment below each quote; here I’ve denoted that text from my original post as “Joe’s original post”) –

Joe’s original post – The big one – being late for or missing an IME is really, really expensive – the payer has to pay the IME doc the $1000 – $2000 anyway, rescheduling an IME incurs more indemnity, attorney, and other expense, and it makes adjusters nuts. The same is true for PT or other medical treatment – although not quite as bad.
Jim’s comment –           Rideshare programs are a great solution for just this point.  The transparencies provided from real time monitoring of rides allows immediate rescheduling if an adverse issue arises.

Joe response to Jim comment – This may still lead to delays in getting the patient to the IME – unless L/U informs you with sufficient time to reschedule – which is likely much more than 5 minutes – I still don’t see how this solves the issue.

Joe’s original post –
WC patients don’t all have smartphones, so figuring out which patients can access L/U is a necessary first step
Jim’s comment –  The benefit of this program is the patient does not utilize commercial applications.  The coordination and monitoring of rides are done through the vendor, such as HOMELINK.

Joe’s original post –  L/U drivers wait a maximum of 5 minutes – this doesn’t work for patients
Jim’s comment –         Through HOMELINK’s processes, every patient is qualified before the coordination of the ride.  Patients with specific needs or limiting diagnoses are coordinated with transports meeting their exact needs.

Joe response to Jim comment – And L/U aren’t able to accommodate most of these patients, supporting my assertion.

Joe’s original post –
L/U drivers aren’t going to go up to the patient’s door and help them get into the vehicle
Jim’s comment –  As mentioned before, these patients would undergo our initial Rideshare assessment and coordinated with transports meeting their exact needs.

Joe response to Jim comment  – And L/U aren’t able to accommodate most of these patients, supporting my assertion.

Joe’s original post –
L/U drivers’ vehicles typically aren’t accessible for patients with limited physical ability
Jim’s comment –  This would be vetted in the assessment process as outlined above.

Joe response to Jim comment – And L/U aren’t able to accommodate most of these patients, supporting my assertion.

Joe’s original post –
Some WC patients aren’t native English speakers, and may require drivers to speak their language
Jim’s comment –  This would be vetted in the assessment process as outlined above.

Joe response to Jim comment – And L/U aren’t able to accommodate most of these patients, supporting my assertion.

Joe’s original post –
If you try to schedule a pickup for a time in the future, L/U typically don’t reach out to drivers until a few minutes before that pickup time – which may well not work for patients in remote areas
Jim’s comment –  Since HOMELINK coordinates the referral, these necessary communications all occur within our patient care teams.

Joe response to Jim comment – This doesn’t refute my core assertion – there’s still a 5 minute rule.

Joe’s original post –
which brings us to location – L/U work really well in metro areas, and not so much in more rural/suburban settings. So, schedulers may find L/U doesn’t have available drivers
•im’s comment –  Rideshare and traditional transportation are not mutually exclusive, Rideshare is a cost savings tool to be used by payers when the situation allows.  In this example, if Rideshare did not have adequate coverage a traditional transport would be coordinated.

Joe response to Jim comment – never said rideshare and traditional WC transport were “mutually exclusive” – far from it.

Joe’s original post –
you can’t talk to a person at L/U – ever. Many WC transportation services involve a three-way call between the scheduler, driver, and patient; this is just not going to happen with the ride-sharing services.
Jim’s comment –  Our relationship and developed workflows allow our patient care coordinates to have open communications with the driver.

Joe response to Jim comment  – This doesn’t refute my assumption – are these calls three way ?

Joe’s original post –
Finally, while the dollars are significant in WC, they really don’t amount to that much compared to other opportunities in Medicaid and Medicare. If you’re L/U, you’re going to focus on those payers and ignore WC.
Jim comment –  If you’re looking at Rideshare organizations making a direct play with WC you are correct.  However, relationships established in this space with companies such as HOMELINK have allowed for payers to capitalize on substantial savings.  We have shown comparative savings (Rideshare vs. traditional transportation) approaching up to 40%.  This is not an exclusive tool in a transportation portfolio, it is another innovative approach to improve patient experience and optimize savings.  Since our rollout in 2015, we have seen Rideshare utilization becoming a larger percentage of our partner’s transportation mix.

Joe response to Jim comment – I asked what percentage of rides are rideshare, Jim responded it is “in the double digits” but would not go beyond that.  Thus, it’s not possible to determine if this “savings” is meaningful or not.  “approaching up to” is marketing-speak – what is the actual percentage saved for what percentage of rides?

The net – HOMELINK is clearly trying to make the best possible use of rideshare, and kudos to the company for it’s efforts. Much of what they’ve done operationally is to deal with the very real challenges inherent in the L/U business model – a model that is NOT going to change to serve a very tiny market. I’m still waiting to see much of an impact – for reasons I note a couple of paragraphs below.

Re my correspondence with the anonymous commenter re OneCall; s/he made several assertions about OCCM’s use of rideshare. However, I emailed s/he back to ask how s/he knew this, and whether s/he is a OCCM person or customer, and never heard back. (note – the commenter is from Jacksonville, FL, which, perhaps coincidentally, is OCCM’s home town). I’m happy to hear from OCCM, as I noted in the post, or from OCCM’s transportation customers, but readers do need to know who you are to properly understand your statements.

Takeaway – Absent any credible data, it is not possible to quantify the impact of rideshare on work comp transportation. Therefore, we’re stuck with assumptions, albeit ones based on experience.

More important takeaway – There have been any number of “innovations” coming into work comp from outside the industry; models based on what works in group health or Medicare, or efforts to adapt technology and business processes to “fix” problems inherent in workers’ comp.

Most have not worked, or if they have, the impact has been minimal. There are a bunch of reasons for this:

• Every state is different, thus vendors must adapt their tools/approaches/business models to fit 51 different sets of requirements.
• Work comp is tiny and getting tinier – It accounts for 1.25% of total US medical spend, and claim counts continue to decline. Thus the payoff for the investment needed to adapt technology/programs/approaches can be vanishingly small  – or even negative – in states that have relatively low work comp spend.
• But (most) vendors must be able to serve customers everywhere – so you are going to lose money in many states and hope you make your profits in the handful of big ones with a lot of comp exposure – California, New York, New Jersey, Texas, Pennsylvania, Illinois, etc.

What does this mean for you?

Innovation in comp will mostly come from inside comp.