Workplace injuries in a changing labor market

Overview

Every year, over a million U.S. workers are injured on the job. These occupational injuries undermine the well-being of workers and the broader economy. The risk of occupational injury varies significantly across occupations, so changes to the occupational mix of the U.S. labor force will likely affect the health of American workers.

The Bureau of Labor Statistics (BLS) has projected that employment growth between 2022 and 2032 will be concentrated in health care and social assistance occupations, as well as computer and information technology occupations. We estimate how these projected changes in the U.S. occupational mix will impact injury risk across the workforce.

We find that employment growth will be concentrated among the highest- and lowest-risk occupations, while jobs with average risk levels will see relatively limited employment gains. In addition, we project that occupational injuries due to violence by persons or animals will likely become more common. This change is largely driven by job growth among home health aides, who have an elevated risk of violent injury in the workplace.

Background

Occupational injuries are a social and economic problem in the United States. An occupational injury is any injury or illness experienced by an individual as a result of their job. In 2022 alone, employers reported 5,486 fatal occupational injuries and 1,483,400 nonfatal injuries and illnesses that caused an employee to miss at least a day of work. These incidents resulted in roughly 75 million lost workdays, imposing an estimated total cost of $167 billion in 2022.

Occupational injuries have consequences beyond lost productivity. The previous estimate includes medical costs of $37.6 billion and administrative costs of $54 billion, in addition to $50.7 billion in lost wages. Furthermore, occupational injuries are associated with negative mental health outcomes for injured workers, who are more likely to show symptoms of severe psychological distress, including anxiety and depression. Higher occupational injury rates are also associated with an increased prevalence of substance use disorders, particularly opioid use disorder.

Not all jobs are made equal: the risk of workplace injury varies significantly across occupations. Between 2021 and 2022, the annual incidence of nonfatal occupational injuries for the highest-risk occupation (law enforcement) was 70.4 injuries per 1,000 full-time workers. Conversely, the lowest-risk occupation (sales representatives) recorded only 0.5 injuries per 1,000 full-time workers—0.7% of the injury rate for law enforcement. With so much variation in injury risk across different occupations, the prevalence and type of occupational injury in each sector has important implications for the health of its workforce.

The U.S. economy is changing in ways that will alter the occupational mix and, therefore, the incidence of occupational injury across the workforce. Service industries have grown significantly since the middle of the 20th century, dominating employment in the private sector, while employment in goods-producing industries has stagnated in absolute terms and has declined as a share of the overall workforce. Employment in manufacturing has declined by 34% since its peak in 1979, while employment in agriculture fell by 17% between 2000 and 2021.

Taking these economic trends into account, BLS has estimated how total employment in each occupation will change between 2022 and 2032. They project that overall employment will increase by 4.7 million jobs. This growth will be driven by the health care and computer sectors. Home health aides are the job with the largest projected increase in employment and will be the largest occupation in the economy by 2032. The second-fastest growing occupation is software engineers. Job losses will be concentrated in retail and administrative support occupations.

In this paper, we investigate the implications of changing occupational structures for workplace injury risk in the U.S. labor force. Linking employment projections with occupational injury data from BLS, we find that employment growth will be concentrated in both the highest-risk occupations (e.g., home health aides) and the lowest-risk occupations (e.g., computer occupations). Employment growth in occupations with average risk levels, such as food and beverage service workers, will be comparatively limited.

We then estimate how the changing occupational composition of the workforce will affect aggregate injury risk across all employees, holding constant each occupation’s injury risk at 2022 levels. We find that compositional changes alone will result in a modest increase in overall injury risk for American workers. However, we note that injury risks will likely also change within occupations over the intervening decade in ways that we cannot project, which will also influence the overall injury risk in 2032.

Finally, we break out our results by the underlying causes of occupational injury. We find that the risk of violent injuries is projected to increase by almost 4% across the U.S. workforce. This change is largely driven by job growth among home health aides. Home health aides are among the occupations with the highest rates of workplace injuries. Home health aides who serve people with severe mental illnesses, dementia, or work in households experiencing high levels of stress are regularly subject to violence from patients and their families. Home health aides are also projected by BLS to be among the fastest growing occupations in the coming decades.

Methods

We construct a dataset describing the following for each occupation:

• Projected changes in total employment between 2022 and 2032, from the BLS Occupational Outlook Handbook
• The incidence of occupational injuries, from the BLS Survey of Occupational Injuries and Illnesses (SOII)
• Employee demographics such as gender, race, and ethnicity, from the 2022 American Community Survey (ACS)

Occupations are classified based on the 2018 Standard Occupational Classification (SOC) system. BLS includes both illnesses and injuries in these data; we group the two under the general umbrella of “occupational injuries” for simplicity.

Our analysis focuses on the prevalence of occupational injuries that require an employee to take at least one day away from work. BLS also collects information about fatal occupational injuries. However, fatal events are rare—they account for less than 0.4% of all occupational injuries—so those occupation-level data are typically “noisy” and can vary significantly from year to year. For that reason, we focus on nonfatal injuries as our primary outcome.

We categorize occupations into four risk quartiles based on the rate of all occupational injuries. The first quartile represents the lowest-risk occupations, while the fourth quartile represents the highest-risk occupations. The quartiles are not weighted by employment: each quartile represents an equal number of occupations, not an equal number of employees. Table 1 shows the average injury rate by quartile, as well as the three largest occupations in each quartile. The highest-risk quartile has a mean injury rate of 25.6 annual injuries per 1,000 full-time employees (FTE), more than 14 times the mean injury rate in the first quartile (1.8 per 1,000 FTE). 

The causes of these injuries vary across occupations. Common causes include overexertion and bodily reaction (e.g., repetitive strain injuries); falls, slips, and trips (e.g., falling downstairs); and contact with objects or equipment (e.g., being hit by machinery). However, many occupations include unique risks: for example, health care workers experience elevated rates of violent assault in the workplace.

An interactive version of our dataset is available in Table 2. Readers can search individual occupations and sort the data by projected employment growth, 2022 total employment, injury rate, median salary, or employee demographics.

Using the combined employment, injury, and demographic data, we investigate how compositional changes in the American workforce will affect occupational injury risk. Our findings are presented below.

High-risk jobs are projected to grow the fastest

Occupational growth will be the largest among the highest-risk occupations like home health aides and motor vehicle operators (see Figure 1). Occupations in the riskiest quartile, with injury rates ranging from 16.4 to 70.4 injuries per 1,000 FTE, are projected to grow by 5.5% on average.

The lowest-risk jobs are also expected to grow substantially, with occupations in the lowest-risk quartile projected to grow by an average of 3.8%. The occupations in that category include computer professions and business operations specialists, with injury risks between 0.7 and 2.3 injuries per 1,000 FTE. Occupations with average risk levels (quartiles 2 and 3) are projected to have comparatively low employment growth: 0.4% and 0.9%, respectively.

The percentage changes depicted in Figure 1 represent net changes in employment between 2022 and 2032. Figure 2 disaggregates raw employment gains and losses by occupational risk quartile.

We find that the highest-risk occupations will account for 55% of the net increase in jobs between 2022 and 2032. This trend is driven by significant gains in employment for these high-risk occupations, paired with very few job losses.

In contrast, occupations with average risk levels face offsetting gains and losses, meaning that the net change in employment is small. For example, occupations in the second risk quartile will lose 660,000 jobs while gaining 822,000, resulting in a net gain of 162,000 (or 4% of projected job gains).

Because most net employment gains occur in the highest-risk and lowest-risk jobs, these results suggest there will be an increasing divergence in workforce risk levels across occupational groupings. That is, the future labor force will increasingly be distributed into either high- or low-risk occupations. However, it is unclear what these trends imply for aggregate risk across the workforce. We investigate this question in the following section.

Overall injury risk will increase, and violent injuries will become more common

To determine how occupational injury risks may change across the entire workforce, we compare estimated overall risk levels in 2022 and 2032. We estimate the overall risk levels in both years by taking the average of the injury rate across all occupations, first weighted by their total employment in 2022 and again weighted by their projected total employment in 2032. This approach assumes that each occupation’s injury rate will not change between 2022 and 2032. 

BLS disaggregates occupational injury rates by the cause of the injury, categorizing the causes into seven distinct categories (detailed in the Technical Appendix). We use those data to separately estimate how the prevalence of injuries by cause will evolve. Figure 3 illustrates the estimated changes in injury rates, both for all injuries and separately for each cause.

Figure 3 indicates that the net increase in the risk of occupational injury will be approximately 0.9%. Putting this finding in perspective, if one applies our estimate to the 2022 number of occupational injuries (1.483 million) the result is an increase of over 13,000 injuries that would, in turn, cost about $1.5 billion using National Safety Council estimates.

The largest increase in occupational injury by cause will be those associated with workplace violence (more specifically, injuries caused by “violence and other injury by persons or animals”). In recent years, the workplace violence rate has been 0.7 injuries per 1,000 full-time workers. We estimate that there will be a 3.7% increase in the rate of injuries stemming from workplace violence. Most of that increase is due to either intentional violence by people or animal-related violence (see Figure 4). Acts of intentional violence will represent over half of the additional violent injuries in 2032.

These results assume that injury rates will not change for each occupation between 2022 and 2032. However, recent evidence demonstrates that injury rates have changed for some occupations at least. Trends are mixed: for example, injury rates have declined for construction trades, while health care workers have seen increased rates of injury since 2011. Within an occupation, these trends may also look different for each injury type. As a theoretical example, health care workers may fall less often, but they could also experience workplace violence more often. Furthermore, the COVID-19 pandemic impacted injury rates from 2021 to 2022, both by increasing the number of workplace exposures to viruses and by decreasing workplace injuries for occupations where working from home became the norm. Finally, there is significant uncertainty regarding future factors, such as new technology, that may influence injury rates within different occupations. Given these considerations, we are not able to project changes to injury rates within occupations. Instead, our results focus solely on changes in injury risk due to compositional shifts in the American workforce (see the Technical Appendix for more details).

Growth in violent injuries will impact direct care workers

A substantial share of the overall increase in violent injuries will be driven by growth in the occupational group encompassing home health and personal care aides, nursing assistants, orderlies, and psychiatric aides if no new actions are taken to curb occupational injury rates. Notably, home health and personal care aides account for most of this group (over 72%). We refer to this group as “direct care workers” for simplicity.

We predict that in 2032, there will be 7,155 more violent occupational injuries annually than there were in 2022. Direct care workers account for 45% of that increase: they are predicted to experience 3,199 more violent injuries annually in 2032 compared to 2022.

This finding highlights the unique position occupied by direct care workers in the American labor force. Direct care workers are expected to see significant employment growth, particularly among home health aides. This growth has multiple causes: first, the U.S. population is aging rapidly. Between 2010 and 2020, the population of adults ages 65 and over increased by 15.2 million people (38%). As the number of older adults increases, so will the size of the workforce required for caregiving. In addition, the past decades have seen a shift from nursing homes towards home- and community-based care for older adults, compounding the need for home health aides and other community-based direct care workers. Finally, the supply of informal home caregivers has declined due to the increased participation of women in the formal labor force. As these trends continue, the need for direct care workers will be significant: for example, BLS projects that home health aides will account for one in six new jobs and will be the largest occupation in the U.S. by 2032.

In addition to being the occupational group with the largest projected growth, direct care workers also face elevated risks for occupational injury. With an annual incidence of 29 injuries per 1,000 FTE from 2021 to 2022, direct care workers were injured almost 2.5 times more often than the rest of the workforce. Compared to the other occupations, direct care workers were disproportionately exposed to violent injuries (5.5 times as often), overexertion (2.4 times as often), and falls (1.5 times as often).¹

Furthermore, our analysis shows that direct care workers are overwhelmingly women (86.2%) and non-white (62.5%). They are also low-paid, with a median salary of approximately $31,000, and only 40.3% of direct care workers report receiving employer-sponsored health insurance through their own or a family member’s job.

Figure 5 shows how direct care workers compare to the other 94 occupational groups in our sample along the dimensions listed above. It demonstrates that direct care workers are outliers across multiple outcomes. Of all occupational groups, they are the:

• 1^st in projected growth,
• 2^nd least white,
• 4^th most female,
• 4^th highest risk,
• 6^th least covered by employer-sponsored insurance,
• 8^th lowest paid,
• and 12^th least English-speaking occupational group.

Given the many social and economic inequities faced by direct care workers, it is particularly concerning that many of the additional injuries resulting from compositional changes in the workforce will affect this group. Reducing injury rates would keep more direct care workers, especially home health aides, at work. It may also reduce turnover in the direct care workforce, which could help address the shortage of direct care workers and potentially benefit clients as well as employees.

Violent injuries disproportionately impact women and Black employees

Which demographic groups might be impacted by the predicted increase in violent injury? We can use 2022 data to understand which employees are currently impacted by violent injury. Unless the composition of high-risk occupations changes dramatically between 2022 and 2032, the population of employees will follow similar demographic trends.

First, we examine the current incidence of violent injuries by gender. While many of the highest-risk occupations are male-dominated (e.g., law enforcement, supervisors of protective service workers), employment in these occupations is comparatively small, and it is not clear how violent injury risks correlate with employee gender in aggregate. We divide occupations into two groups: female-dominated occupations, which employ over 50% women, and male-dominated occupations whose employees are over 50% men. We study average rates of violent injury for each group, disaggregating by the type of violence that resulted in injury. The results are displayed in Figure 6.

Male- and female-dominated occupations have similar rates of unintentional and animal violence.² However, we find that female-dominated occupations have higher rates of injuries due to intentional violence. The female-dominated occupations with the highest intentional injury rates are other educational instruction and library occupations (2.96 injuries per 1,000 FTE), direct care workers (2.59 injuries per 1,000 FTE), and health technologists and technicians (1.60 injuries per 1,000 FTE).

We also study how injury rates vary by employee race. In Figure 7, we divide occupations into two groups based on their share of Black employees. The median occupation employs 9.5% Black employees. All occupations with over 9.5% Black employees are placed in the “More Black Employees” group, while the “Fewer Black Employees” group represents occupations with a workforce that is less than 9.5% Black.

Figure 7 shows that occupations with more Black employees have a substantially higher incidence of both intentional and unintentional injuries. Jobs with higher shares of Black workers experience 70% more intentional violence and 140% more unintentional violence than occupations with fewer Black workers. These occupations include law enforcement workers; supervisors of protective service workers; other educational instruction and library occupations; and direct care workers. However, occupations with more Black employees are comparatively less exposed to animal violence.

However, injury rates do not follow similar trends when occupations are divided by Hispanic/Latino employee share. In Figure 8, occupations with more than the median share of Hispanic/Latino employees (14.6%) are placed in the “More Hispanic/Latino Employees” group, while occupations below the median are placed in the “Fewer Hispanic/Latino Employees” group. Injury rates follow very similar trends across both groups, though occupations with more Hispanic/Latino employees have marginally higher rates of unintentional violence and marginally lower rates of animal violence.

Conclusion

We project that compositional changes to the U.S. labor force between 2022 and 2032 will marginally increase the incidence of occupational injuries among American workers. We predict that violent workplace injuries will increase by nearly 4% due to job growth among direct care workers and other service occupations. It is likely that this increase in violent injuries will disproportionately affect female and Black employees, who are overrepresented in jobs with high rates of workplace violence.

Occupational injuries do not occur in a vacuum. Employers can take measures to reduce the incidence of occupational injuries and mitigate harm when they occur. These measures can include updating their business practices as well as their physical work environments. Employers could also benefit from implementing injury prevention programs, which may help to attract and retain employees in sectors currently facing shortages such as home health care. Given the projected growth of high-risk jobs and occupations with high rates of violence, it is important that employers take measures to keep American workers safe on the job.