Articles & Blogs

Preposterous Presumptions: COVID-19 Workers’ Comp Presumption Laws Trivialize Science, Accomplish Little for Workers, and Unnecessarily Burden Employers

January 26, 2023
April 17, 2020

Legislatures around the country are rushing to enact state statutes declaring that COVID-19 is presumptively related to various classes of occupations, triggering employer obligation to pay workers’ compensation benefits to certain employees.  These statutes have no basis in science, will typically provide workers with less money than other programs already in place, and place an extraordinary burden on struggling businesses and municipalities.

Some states have passed “irrebuttable presumption” laws that declare people in whole categories of work to have contracted the virus on the job, and the rest of the states will likely follow in lockstep.  These presumption statutes are both overinclusive and underinclusive because they have no basis in science and vary from state to state, as if the virus effects people in one state differently than in neighboring states.  Irrebuttable presumptions essentially slam the door on any employer effort to distinguish one case from the next or to apply a scientific methodology for determining the causation of a workers’ infection.

This virus scares the wits out of people.  The unpredictability of the course of the disease and constant worry about catching the virus have led some people to prudent caution and others to near hysteria.   Not surprisingly, in such a crisis, leaders feel pressured to “do something.”  Those in our federal and state legislatures are under political pressure to demonstrate leadership and sincerely want to take action to help those hit the hardest. But the rush to act has virtually eliminated any serious consideration of the science of this disease and obscured the fact that most people who are sick and disabled by the virus, are already better covered under programs outside of state workers’ compensation.

States have taken nearly arbitrary lists of occupations or job titles and thrown them into presumption legislation.  In many states, the presumption applies to only first responders, police and firefighters.  Other states are adding or substituting health care workers, nurses and physicians.  Yet others have singled out delivery workers, cashiers and retail clerks.  Still others include transit workers, home health care attendants and cleaning service employees.  Some require proof of some element of job exposure to the virus, others do not.  One even grants the presumption to members of the media. Some grant the presumption to volunteers; some grant the presumption retroactively to cover employees exposed before the effective date of the presumption and few attempt to coordinate benefits with other programs.  The inconsistency between the laws of various states is a testament to the fact that none are based on hard, proven, science.

The use of broad occupational categories may be overinclusive and unfair.  For example, if a state statute deems COVID-19 to be work-related for all municipal employees, an office-based deeds recorder would be covered, even if the employee had no direct exposure to people who might have the virus, just the same as an EMT running continuous routes to the hospital.  Similarly, a firefighter assigned as a first responder may have repeated and intense exposure to seriously ill people, but a firefighter assigned to the arson investigation squad may have had little interaction with any member of the public, sick or not.  Broad occupational categories are a poor substitute for evidence of exposure.

At the same time, the use of broad occupational categories can be underinclusive and unfair.  Whereas doctors and nurses typically have training, supervision, access to protective equipment, access to handwashing sinks, access to gloves and gowns, city bus drivers have perhaps more day to day contact with the public but have virtually no training, equipment, or ability to wash their hands or change their clothes during a shift.  Why provide the presumption to the protected and trained medical personnel but deny the presumption to the untrained and unprotected driver?   As noted below, of the very limited surveys of which occupations have the virus most often, an argument can be made that city bus drivers, and not medical personnel, should be included in a presumption, if a presumption is to be allowed at all.

The problem is that these presumption statutes are based only on “gut feelings” that people in certain occupations have a higher chance of getting the virus.  Such hunches may not be true for two reasons.  First, this form of reasoning ignores the fact that people who are not in the designated occupations also get the virus, perhaps more frequently than those in the specified occupational group.  Second, this thinking ignores other factors that might be at play, such as worker exposures off the job.  All sorts of confounding factors might be at play, ranging from geography to socioeconomics, that could make a certain group of workers look more exposed.

For statutory presumptions to have any logical validity, studies must be conducted on the incidence of COVID-19 by occupation, with controls for factors like socioeconomics, worker age, geography, dose/response relationship, gender, smoking history, population density, and the presence or absence of other confounding factors.

Step one in the process of determining whether a presumption of causation is justified for any occupation would be gathering data.  Ideally, we would want to know the occupation of every person with the virus.  We could then analyze the data and determine if people in one occupation have a statistically significant increase in the incidence of the virus.  But this will be impossible for many reasons.  First, somewhere between 15% and 70% of all people with the virus have no symptoms, and therefore, were never diagnosed, and we will be unable to get a full count of people who had the virus.  Without knowing the number of people who had the virus, we obviously can’t parse out the number of people in each occupation who had the virus.  

Furthermore, many people had symptoms, but recovered before ever being tested.  To date, only about 3 million Americans, or less than 1% of the population, have been tested.  We have no way to determine the occupations of the millions of people who were either symptom-free or recovered from the virus without ever being tested.  

Moreover, even those who tested positive or were treated for the virus, including those who died from the virus, were not necessarily queried extensively as to their job titles and duties.  For example, news reports seem to suggest that a lot of death occurred among the elderly nursing home population.  For some patients, things were too hectic to even verify that they had the virus at the time of death, let alone catalog the nature of their exposures before getting sick.

It seems that we are therefore handicapped in our ability to develop meaningful data on the incidence of COVID-19 by occupation.  Due to “detection bias”, meaning that only people who were tested can be included in a survey, studies will be limited.  To date, there is no solid and reliable model to determine whether one group, whether defined by occupation, race, age, or gender, is more afflicted than any other.  Theories abound that minorities, city dwellers, older people, smokers, or obese people are most affected, but it will take years before anyone can prove, with any level of certainty, that the impact of the disease is not simply random, let alone attributable to designated occupations.

It is interesting to see the speculation in each state about which occupational groups should be declared to have presumptively causative exposure to COVID-19, In Alaska, for example, only first responders and “health care workers” are included in the presumption.  While in Pennsylvania, the laundry list includes an:

"Individual employed by a life-sustaining business or occupation." Front-line employees and other individuals employed by or under contract with a life-sustaining business or entity who work during the declaration of a disaster emergency or a public health emergency. The term includes, but not limited to:

(1) First responders, including law enforcement officers, firemen, emergency medical technicians and other individuals who are considered to be first responders.

(2) Correction officers.

(3) Emergency services dispatchers.

(4) Ambulance drivers.

(5) Retail workers, including restaurant, food services and grocery store workers, cashiers and other support staff.

(6) Food and agriculture workers.

(7) Medical, health care and public health workers, including doctors, nursing professionals, physician assistants and paramedics, and other support staff.

(8) Pharmacists and any cashiers and other pharmacy support staff.

(9) Home healthcare workers.

(10) Public utility workers, including workers engaged in providing telecommunications, energy, water and wastewater services and public works.

(11) Any employee of state or local government.

(12) Trash collectors.

(13) Warehouse workers.

(14) Any other individual employed by a life-sustaining business or occupation who is required to work during the…(emergency)…”

Left off the list in Pennsylvania are clergy, chaplains, delivery people, taxi and Uber drivers, teachers, school custodians, cleaning service employees, domestic workers, gas station attendants, auto mechanics, social service providers, childcare providers, etc.  

On the other hand, if the list is expanded to include any person who may have been exposed on the job to the virus, it is likely that the entire employed population of America would have to be included, and as noted below, maybe the entire employed population should be included in wage loss benefits due to COVID-19.  

Despite the data limitations on accurately determining the occupations of those most infected, there are anecdotal reports from trade associations, municipalities, and media outlets that offer snapshots of some sort of data, but nothing near the level of consistency and reliability to base an analysis on.  One foreign and preliminary source suggests that 85% of all COVID-19 cases are completely untraceable to any occupation at all [1], but many sources have tried to highlight incidence figures in certain occupations.

For example, Alaska felt the need to include “first responders”, but not bus drivers.  First responder is a broad category, that includes, at one end of the spectrum, an EMT who transported 100 COVID-19 patients to a hospital, and at the other end of the spectrum, a police precinct captain assigned to desk duty.  But for the sake of argument, let’s assume that all police and firefighters are first responders and all were exposed to COVID-19.

According to one police industry website, there were 29 police officers in America lost to COVID-19 [2].    According to a firefighter industry website, there were 13 firefighter deaths in America attributable to COVID-19 [3].  There are 1,115,000 firefighters in America [4].  There are about 1,001,000 police officers in the United States [5]. Therefore, out of 2,226,000 police and firefighters, 42 such people have died in the service of their communities.  

For the sake of comparison, so far, about 61 transit workers have died of COVID- 19 in New York City alone.

42 teachers and school staff have died as a result of COVID-19, just in the public schools of New York City.

But despite these early statistics, states are rushing to grant comp to police and firefighters, ignoring some occupations that might have more deaths, but less public sympathy, like teachers and bus drivers.  

Enacting statutory presumptions is just a way for legislatures to appear to be active and concerned, while merely pretending to know which, if any, industry group is more impacted than any other.  Picking out occupations to assign a presumption is not only ill-conceived, it is unnecessary.

The goal of workers’ compensation programs is to provide a replacement for wages lost during absence from employment due to work-related injury or disease.  In most states, these wage replacement (“indemnity”) benefits are paid at around 66% of normal salary, but subject to maximum amounts.  In many states, the maximum amount paid under workers’ comp is capped at around $1000 per week.  The range is a low of $468 in Mississippi to a high of over $1600 in Iowa.

But most workers who lost time from employment due to COVID-19 would get a lot more money under existing federal laws than they would under these contrived presumption laws in the states.  Under the CARES Act, signed into law on March 27, 2020, three new alternate routes to compensation were established:  enhanced sick leave, enhanced FMLA, and enhanced unemployment.  These new federal programs provide money to effected workers whether their COVID-19 was related to their jobs or not.

Under the new sick leave law, a worker can get up to 10 days off from work at 2/3’s of regular pay, without being subject to the state maximums on workers’ compensation.  For example, a worker in Pennsylvania earning $120,000 a year would get only about $1100 a week in comp, due to that state’s maximum amount limits.  Under federal sick leave, the employee would get $1538.

Under the new enhanced FMLA, a worker with the virus can get up to 16 weeks of benefits at a maximum of $200 per day.  Using our same hypothetical PA worker making $120,000 a year, the worker would get $1538 in the first two weeks, followed by 14 weeks at $1000 a week, for a total over the four-month period equal to state workers’ compensation.

Under the new unemployment law, a worker can, for the first time, get unemployment even if the worker is sick and unable and unavailable for work.  Each state has its own formulas and maximums for unemployment compensation, but the average is a maximum of about $500 per week.  However, the CARES Act enhanced the state benefits by $600 per week, meaning that the new average is about $1100 per week.  Again, using our $120,000 per year hypothetical Pennsylvania employee, the worker would get $1197 per week for 16 weeks under unemployment, which is more than the worker would get under comp.

Another huge benefit of using existing federal compensation programs instead of workers’ comp payments is that for the most part, any payments noted above are either made entirely by the government or returned to the employer by the government via a dollar for dollar tax credit.  In essence, these programs get most workers the same or more money, and at the same time, cost the employer virtually nothing.  Comp, on the other hand, costs the employers dearly in annual premiums, administrative fees, and retrospective premium assessments.  The comp process also encompasses delays, bureaucracy, and administrative expenses.

So why are legislatures rushing to weave statutory presumptions out of thin air to force people into the workers’ compensation system, instead of allowing workers to get more money and employers to pay less money by admitting that no one knows in a given case, and no one can know whether a particular worker’s virus was in fact and truth caused by employment?

I can think of only a few reasons, none of them palatable.

One is that in the federal plans, attorneys are not necessary, and therefore, attorneys cannot get fees out of the workers’ benefits.  In comp, attorneys typically siphon off 20% of the benefits that would otherwise go to workers.  Lawyers have a big interest in using the comp and litigation systems instead of allowing the benefits to flow to workers free and unencumbered.

Another is that legislatures are under political pressure to do something – anything – that appears to alleviate the hardships placed on constituents.  Workers vote.  Companies do not.  By shifting liability to companies, legislators might seem more aligned with voters.

Legislatures are also bombarded by lobbyists, including union lobbyists.  They can be duped by influential pitches and one-sided proposals, especially when the outcome is likely to help solidify voter and financial support.

Another reason is that states tend to follow the lead of other states like sheep.  If State A does something, there is enormous pressure on State B to follow suit, to prevent the appearance of acting slowly or being one-upped by another state in the eyes of constituents.

Finally, legislators are humans, and like all humans, are genetically programmed to see patterns in life, even if they don’t exist.  Like all humans, they want to feel in control and want to feel like they understand the workings of the world.  It is natural to want to help, empathize, and form beliefs that align with these goals.  It is entirely unnatural to apply a scientific methodology to a problem, reach conclusions only after adequate data has been collected and analyzed, and to be satisfied with the notion that we humans are simply incapable of knowing the cause of everything.

Not only are presumptions statutes illogical and unnecessary, they may also be illegal.   Constitutional challenges should be mounted against irrebuttable presumptions of this ilk, and against the retroactive application of a new conclusive presumption in cases that arose before the passage of the presumption legislation.  

Rather than knee-jerk and lock-step enactment of scientifically baseless presumption statutes that do no particular good for workers, and substantial harm to employers, legislatures should fade into the sidelines and rely, for once, on existing federal legislation that seems to make sense. [6]






[6] If a legislature is compelled to enact a presumption, it should, at the same time, enact provisions requiring the worker to apply for and use the federal benefits first, and to offset any workers’ compensation benefits by the amount received in such benefits.  Further, health care insurers paid for or sponsored by employers should be prohibited from attempting to collect reimbursement (“subrogation”) from the workers’ comp carriers.  Otherwise employers and their comp carriers will bear the brunt of millions of dollars in unanticipated medical expenses that, without the legislative presumption, would certainly have been paid by the health insurer.