This year represents a significant milestone in the history of the workers compensation (WC) system—the
100th Anniversary of the National Council on Compensation Insurance (NCCI). NCCI serves as the nation’s most comprehensive source for WC data, insights, and solutions. As NCCI celebrates 100 years of excellence, let’s reflect on how WC ratemaking has changed over the past century.
In 1923, NCCI provided services for 69 WC insurers.1 Today, that number is nearly 900.2 The WC system has continually adapted to the changing environment, including states updating the name from workmen’s compensation to workers compensation.
While the challenges currently facing the WC system may seem unique, the themes are quite reminiscent of historical challenges the industry has faced—inflation, changing benefit levels, data availability, and catastrophic events. There is much to be learned from the past to help us navigate today’s challenges. This article highlights major changes over the years impacting key ratemaking components including collecting data, choice of exposure base, experience period, trends, cost drivers, and the residual market.
Before any ratemaking can take place, it is necessary to have data. In fact, the first task of the National Council on Workmen’s Compensation Insurance in the 1920 revision of rates was to “collect and compile experience for ratemaking purposes.”3 At that time, Schedules W and Z, based on Unit Statistical Data were the primary sources for ratemaking. The lag in data availability was noted as a ratemaking challenge. Since that time, significant strides have been made in automating the manual delivery of data. However, as technology advances, and expectations rise, data availability lags remain a challenge—in part due to necessary validation efforts to ensure data is appropriate for ratemaking purposes. Technology enabling increased transactional data reporting is continually improving this process.
In the 1970s, aggregate financial data (i.e. annual premium and loss data) became the primary data source for state-level ratemaking. This served multiple purposes, including providing fresher, timelier, and verifiable data facilitating the ability to recognize the significant loss development that occurs at later reports in the WC line of business. The long-tailed nature of WC is, in part, due to the lifetime medical benefits provided. In 1988, NCCI recognized the need for even more data, expanding its policy and calendar year calls to include an additional seven years of data.
In the early 2000s, the Large Loss and Catastrophe Call was implemented to support NCCI’s new aggregate large loss procedure. While the aggregate Financial Call data serves many purposes, more granular data facilitated this methodology enhancement. Also, 2008 marked the beginning of another call expansion, adding 10 more individual years of data to the calls.
An important consideration in ratemaking is the choice of exposure base. Looking as far back as 1919, the industry noted payroll as the exposure base. In the 1960s, in most states, the exposure base was payroll limited to an average of $300 per week—referred to as “limited payroll.” Amid the inflationary environment in the mid-1970s, it was apparent that many employees were earning more than that weekly limit. As a result, limited payroll became a proxy for headcount. For example, an employee working 40 hours per week was charged the same premium as an employee working 60 hours per week, despite the latter having greater exposure to injury. This topic remains very relevant, given recent employment level changes.
In the mid-1970s, the WC exposure base changed to total payroll with a few exceptions for high-salary jobs. This is still predominantly the exposure base used today.
The chart below shows total wages and salaries by year, beginning in 1930.4 Wages and salaries steadily increased since that time, indicative of the strength of the US economy with a few exceptions. Notably, in 2008 and 2009 (during the Great Recession) there was a decline in employment and wages across many industries, particularly contracting.
Ratemaking is prospective in nature. The experience period represents the number of years of historical premium and loss data used as the basis for a rate/loss cost filing. The experience period used by NCCI varies by state. With less data available 100 years ago, large states relied on a single policy year of data. By the 1960s, NCCI used the most recent two policy years as the basis of rate indications, in conjunction with an adjustment factor based on the most recent calendar year. In the 1970s, this again changed, and equal weight was given to policy year data and calendar year data.
Over the years, the experience period continued to evolve—moving to three policy years, then to four and five in some instances. These changes recognized state variation and the need to achieve a “proper balance between responsiveness and stability,”5 as required by Actuarial Standards of Practice.
Trend factors reflect anticipated changes in the amount of indemnity and medical benefits compared with anticipated changes in the amount of workers’ wages over time. Trend factors may be thought of as a measure of inflation between the time of the historical data on which the rate filing is based and when the proposed rates will be in effect.
While mentions of trend can be found from the 1930s, it is not until the 1970s that trend procedures are documented in the actuarial literature. NCCI first mentioned its use of a loss ratio trending procedure in 1985. Prior to 1990, NCCI relied on a linear-trend methodology to estimate future claim frequencies and severities. In 1990, NCCI began to employ an exponential-trend approach—a methodology more closely aligned with that used in other liability lines. NCCI continually reviews its trend methodologies, given the impact the trend component may have on a jurisdiction’s annual rate/loss cost filing.
Frequency trends and severity trends are components of loss ratio trends. These are also reviewed during the ratemaking process. The chart below shows incidence rates of nonfatal occupational injuries and illnesses since 1975.6 Incidence rates are a proxy for claim frequency. The long-term frequency declines since at least the 1990s may be partially attributed to increased workplace safety, mix shifts toward safer industries, and advancements in technology. Declines in claim frequency are a driver of the recent rate/loss cost level decreases.
Medical costs are
top of mind for insurers today, as they were in the 1970s and 1980s. During this period of high inflation, medical inflation rates hit all-time highs. High medical costs drove large rate level increases in many states. At that time, some states saw multiple rate level increases in the same year, in response to steeply increasing medical inflation. In 1985, for example, 38 of the 43 NCCI states supported and approved rate level increases—18 of which were in the double digits.7 Along with inflation, the 1980s and 1990s saw major reforms regarding compensability and benefit levels. Many states implemented medical fee schedules as a cost-containment mechanism to address rising medical costs.
The chart below shows the Consumer Price Index for All Urban Consumers (CPI-U). Changes in the CPI are a measure of general inflation.8 Inflationary increases were steep in the 1980s and recently showed similar year-over-year changes. However, medical inflation outpaced them, as measured by the Personal Health Care Index shown in orange on the chart below.9 Medical inflation was well below CPI changes in 2021 and is projected to remain so in 2022; potentially indicative of the efficacy of cost containment measures, such as medical fee schedules and drug formularies.
The residual market posed another challenge for ratemaking during the 1970s and 1980s. Worsening experience and corresponding growth in the residual market contributed to rate level increases that occurred during that time.
In response to a growing number of uninsured risks in the late 1920s, NCCI established the Workers Compensation Insurance Plan to ensure coverage availability for risks considered to be uninsurable. Over time, the market share of the Plan fluctuated—drastically at times.
The residual market share increased rapidly beginning in the mid-1980s, rising from 5.5% in 1984, to 16.2% in 1986, and ultimately reaching its peak of 28.5% in 1992.10 To address this growth, the industry focused on depopulating the residual market, for example establishing assigned risk-to-voluntary market differentials, the implementation of the Assigned Risk Adjustment Program (ARAP), and eliminating premium discounts on policies written in the residual market. These types of programs were successful, as evidenced by today’s manageable residual market share of approximately 6%.
The second task noted in the 1920 National Council on Workmen’s Compensation Insurance revision of rates was to “establish classifications for rate making…” In 1919, there were 1,319 classifications leveraged from employer’s liability insurance, and they were occupationally based.12 Through many iterations, the number of classifications reduced to 1,000, then to 700 in the 1970s, and further reduced in the 1990s to approximately 600. Today, there are approximately 550 class codes.13 The reduction in the number of classifications improved the credibility and accuracy of the final rates by minimizing the need to distribute experience among so many classes.
While the number of classifications decreased, the number of industry groupings, to which the class codes are assigned, increased. The three original industry groups were Manufacturing, Contracting, and All Other. The number of groups expanded to five in the early 1990s when Goods & Services and Office & Clerical became industry groups. The addition of the Office & Clerical industry group recognized the growth in this industry—a comparatively high-payroll/low-premium and low-frequency industry group. This Office & Clerical industry group is the largest in terms of payroll.
Open competition in WC in the 1980s and 1990s brought renewed interest in pricing. The transition from rates to loss costs occurred in many jurisdictions, along with an increased popularity of offering WC coverage on high-deductible policies (i.e., those more than $100K in premium).
Ratemaking procedures, which were similar among WC rating bureaus, began diverging in the 1990s—recognizing the widely varying experience, wage-levels, development patterns, trends, and law-evaluation methodologies that existed across states. The need for state-specific considerations in ratemaking is not a new concept. In fact, it was mentioned as early as 1919.14 Today, we continue to recognize both jurisdiction-specific needs and the benefits of collaborating across WC rating bureaus to review methodology.
It is said that to know where you are going, you must know where you’ve been. Since NCCI’s inception, with an actuarial department of seven employees, the role of the ratemaking staff remained the same—to ensure rate adequacy. NCCI will continue to seek input and refine its methodologies. This includes actively reviewing topics such as alternative ratemaking methodologies, disease loads, large loss limiting methodologies, excess loss factors, and loss development by injury type tail factors, and standard wage distributions.
What Barber observed in the 1930s, remains true today—ratemaking is constantly evolving.15 As NCCI actuaries, we are committed to this evolution. Over the last century, NCCI adjusted and enhanced its ratemaking methodologies to adapt to changes in the labor market, economy, and legislative landscapes, while incorporating data and analytics into the process. The next 100 years of ratemaking will require the same adaptability if we are to continue to foster a healthy WC system for many years to come.
Blanco, Robert G., “Workers Compensation Ratemaking in the Early 1990s,” Unpublished Paper
Feldblum, Sholom, “Workers’ Compensation Ratemaking,”
Casualty Actuarial Society, 1993, www.casact.org/sites/default/files/database/studynotes_feldblum5.pdf
Gillam, William R. and Richard H. Snader, “Fundamentals of Individual Risk Rating,”
Casualty Actuarial Society, 1992,
Kallop, Roy H., “A Current Look at Workers’ Compensation Ratemaking,”
Casualty Actuarial Society, 1975,
Marshall, Ralph M., “Workmen’s Compensation Insurance Ratemaking,” Ralph M. Marshall,
Casualty Actuarial Society, 1961 Revision,
Proceedings of The Casualty Actuarial and Statistical Society of America, 1919-1920, Volume VI-NUMBERS 13,14,
This article is provided solely as a reference tool to be used for informational purposes only. The information in this article shall not be construed or interpreted as providing legal or any other advice. Use of this article for any purpose other than as set forth herein is strictly prohibited.