Friend and colleague Peter Rousmaniere recently attended the NCCI conference and was kind enough to provide a comprehensive report. Here it is, and thank Peter when you see him.
How big is the workers compensation sector?
Dennis Mealy, chief actuary, gave his annual state of the line presentation. He started off by estimating the entire size of the insured market — $88 billion in 2007, compared with $59 billion ten years ago in 1998.
The growth tracked very closely to steady trends in no-farm wages and salaries. Accounted for are private carrier premiums, large deductibles, state funds. The cyclical nature of premiums was accounted for by an entry called “adjustment for premium and price changes. I don’t really understand the components. Since the monopolistic state funds such as OH and WA are not included, it is fair to say that total premium (including the value of large deductibles) is higher than $90B.
This of course leaves out self-insureds. I would venture to guess that the self insured market has an ** equivalent premium** of one third the size of the insured market, which would be $30B for 2007. This brings the total to $120B. This is to me a very, very large figure, far more than what I have been carrying around in my head. I’ll use the amount of $100 Billion, but would not object strenuously to a higher figure.
John Burton, the dean of workers comp researchers, has estimated that in 2005 (I think, perhaps in 2006) total benefits paid nationwide was about $55 Billion. The NCCI did not estimate total losses. However, its estimates of total losses, should it make one comparable to Burton’s would be higher because Burton’s estimate is on a cash basis, not an actual estimate of 2007 claims. NCCI would estimate losses on an incurred basis plus IBNR — incurred but not reported, i.e. using inflation adjustments for future payouts – and its figure would be higher. So I would say that total losses in WC for 2007, on an incurred basis, are in excess of $60 – $65 Billion.
Frequency trends
Both the NCCI and the Bureau of Labor Statistics report that the frequency of claims or injuries have been declining by 3 or 4% for many years, since about 1991. The reduction in frequency has been across all major sectors. NCCI doesn’t have an explanation and in the past I have not read an explanation. Personally, I think it is a function of steady re-engineering of the workplace which has boosted productivity and safety improvements have ridden on the coattails. Note that the reduction is in frequency. As the employed workforce increases in size (2% per year?) the absolute number of claims may be dropping at a lower rate.
As noted below, losses per claim are rising at a high rate, eclipsing the flat or declining number of injuries, thus we are in a modestly growing industry dollar wise.
Premium trends
A simple, partial way to measure premium trends is to track “rate/loss cost” approvals over time. These are figures produced by various workers comp bureaus independent of NCCI and NCCI’s own figures for states it directly covers. Mealy grouped all years since 1990 in four phases: 1990-1993, plus 36%; 1994 – 1999, minus 28%; 2000-2003, plus 17%; 2004 -2008 (est.) minus 24%.
Because of rate deviations, below or above the premium figures suggested by loss costs, the path of premiums has been more volatile. For instance, in 2004 – 2006, premiums stayed high while loss costs declined. This produced very good returns to insurers in recent years, on both an accident year and calendar year basis (don’t ask for an explanation of the difference, it will take too long).
There is a widespread assumption, mentioned by Mealy among many others, that premiums are heading down. But, as I will note in a moment, loss cost estimates may be flat or perhaps even heading up, even as competitive premiums decline in the immediate future. The consensus is that premiums will start heading up in about 2010.
Premiums as experienced by the insureds presumably are heading down for two reasons, neither of which is NOT higher interest rates, that old chestnut which is often exclusively used among insurers to explain changes in the insurance cycle. (Sometimes it explains a lot, sometimes little.) One valid reason is that the capital allocated to workers comp is high. The other is that reserve deficiency, which was a very scary $21 billion in 2001, has declined to $2 billion in 2007. Total reserves among insurers today is about $100 billion. A reserve deficiency of $2 billion is thus peanuts. Insurers as a group can feel in strong financial shape.
Plus, combined loss ratios are good – 92 in 2006, 96 in 2007, and return on equity relatively high – 14% in 2006, 12% in 2007. (I say relatively high, because only one other time it has been 12% of higher since 1987, and the industry’s cost of capital as estimated by NCCI is slightly above 10%.) Note: return on equity in the California market in 2005 – 2006 probably averaged in excess of 20%.
Indemnity and medical cost trends
Mealy says that indemnity claims cost have been “quite well behaved.” They rose below wage inflation in 2004 and 2005. In 2006 and 2008 they rose above wage inflation – 4% in 2007 in indemnity, vs, 3.3% in wage inflation. The average indemnity cost of a lost time claim in 2007 was $20,000, compared to about $15,000 in 2000.
However, medical expenses have been growing substantially higher than medical cost inflation – between 50% and 100% higher. In 2007, medical costs rose 6% vs. 4.4% for medical cost inflation. The medical costs of a lost time claim in 2007 was about $25,000, compared to about $15,000 in 2000.
This means that the total cost of a lost time claim in 2007 was about $45,000. Claims costs are rising at a faster rate that frequency is declining. Therefore we are in a modestly growing industry.
From Mealy’s data and other data presented at the conference, it appears that in the late 1990s and the first few years of this decade – say 1997 through 2001 – medical costs really surged. This applies to non drug and drug costs, and to surgery frequency, diagnostic tests, and PT. This is when medical costs really took off. The NCCI people have not talked with the medical community to find out what happened. My guess is that during that time, the medical community became more aggressive in treating, partly due to greater resources such as more trained surgeons and more drugs, and partly due to economic pressures within the medical community.
Accordingly, medical costs as a percentage of total costs rose from 46% in 1987 and 53% in 1997 to 59% in 2007.
A note on drug costs
Drug cost increases have flattened or even declined in the past two years. This is due according to NCCI primarily to what they call “going down the formulary” – inducing doctors to prescribe lower cost drugs. This is not going from brand to generic, since virtually all the shift from brand to generic has already taken place in prior years. There is no evidence that drugs themselves are being used less.
Peter Rousmaniere
Peter is preparing an article which will show that prescribed drug use is far more pervasive among injured workers than we implicitly expect.
