When is the study released?
The study findings are generally released in the fall of each even-numbered year, in summary form. A full report, including detailed data and notes on methodology, is released several months later. Oregon has been doing these studies in even-numbered years since 1986.
What is the data source for the study?
We survey insurance regulators and workers' compensation rating bureaus in each of the 50 states plus Washington, D.C., for rate information, as of Jan. 1 of the study year.
What is the Index Rate that is used in the ranking?
Every state's economy is different, which leads to a different mix of industries, occupations, and occupational hazards. We want to factor out this difference by using a comparable mix of risk classifications for each state. The Index Rate is a type of weighted average, using a consistent mix of 50 major risk classifications, across all states. The weights used in the average are the most recent available Oregon payrolls for each class; thus the mix varies slightly from study to study.
Since you weight your averages using Oregon payrolls, doesn’t it just apply to Oregon?
The most important factor is to control for hazard mix. Without a control for this, you don’t have a valid comparison of what comparable employers would pay. There is no other current data series that controls for hazard mix across all 50 states.
Oregon's class mix is actually quite similar to the countrywide mix in the classifications that have the largest weight in the study, so the rankings would change very little if a countrywide mix was used. We have verified this by comparing our findings of relative rates to an independent source of rate comparison data for 27 states that uses countrywide data as a benchmark. The two data series produce very similar results, with the same states shown as higher and lower rates.
The actual average premium for my state is different from the index rate Oregon computes. How come?
The study index rate is comparable across states within each study. There's no intent to produce an actual average rate for each state. Any resemblance to the state's actual average is coincidental.
Simple state averages have valid uses, but they aren't valid for rate comparison across states, which depend on use of a comparable hazard mix. We try to produce the best picture of relative costs.
My state's pure premium rates went down last year, yet its new ranking is higher. How can that be?
When rates in many states are declining, a state with smaller declines than others can see its relative rank go up. The study incorporates additional employer cost factors, including insurer expenses and state administrative assessments. These may trend differently than pure premiums.
Why don't you consider factors like discounts, experience mods, dividends, and so on?
These factors apply to employers based on their individual characteristics, not the state as a whole. The available data are not consistent for all states, nor are they available within the study year.
Why don't you include self-insurers' costs?
States regulate self-insurers differently, and self-insurers don't pay premiums, so their costs aren't reported consistently across states. We focus on employers who are purchasing insurance so we can treat them comparably.
Do you compare rates in relation to states' cost of living?
Consideration of cost of living is implicitly built in, since premium rates are based on $100 units of payroll, and average wages are generally higher in high-cost areas. Comparison to available state-level data shows very little relationship between rates and cost of living.
Are there any other measures of relative cost available?
An alternative measure is published in each study. That measure is each state's index rate expressed as a percentage of the median (middle) state in that year's study. For some purposes this can be more useful than the ranking values alone. In recent years rates have become more closely bunched together, so small differences in the rate index may affect the ranking quite a few places. This makes it harder to track states over time. In the 2016 study there are 21 states within plus or minus 10 percent of the median, indicating a very tight grouping.
Is more detail available on your methodology?
The full-length report contains details of methodology and notes for users. The studies available online can be found at: http://www.oregon.gov/DCBS/reports/Pages/general-wc-system.aspx#wcystem4.