Healthcare and social assistance professions added 693,000 jobs in 2025. Without those gains, the U.S. economy would have lost roughly 570,000 jobs.
At first glance, these numbers suggest that healthcare is a growth engine in an otherwise slowing labor market. But a closer look reveals something more troubling for patients and healthcare professionals.
Here are three worrisome conclusions about the state of American healthcare based on the latest jobs data.
Conclusion 1: Healthcare productivity lags
Productivity, an economic measure of how many goods or services a worker produces, rises when businesses find ways to generate more value without expanding their workforce. This is what allows wages to grow and companies to strengthen their financial performance.
Across much of the U.S. economy, businesses are using generative AI and other technologies to help employees accomplish more. As a result, revenues are rising in nearly all industries without proportional increases in staffing.
Healthcare is the exception. As demand for medical services continues to grow, hospitals and health systems have responded by hiring more clinicians and staff rather than redesigning care to increase productivity.
That dynamic is pervasive in healthcare, compromising household, employer and government budgets. National health expenditures have risen roughly 7% annually while overall economic growth (GDP) and inflation average closer to 3%.
Although many factors contribute, medicine’s productivity gap can be traced primarily to three systemic failures:
- Resistance to AI technology. In industries like software, finance and customer service, generative AI is increasing worker output for complex, high-value tasks. In healthcare, however, GenAI has been used primarily for documentation and billing support. But that’s not where most healthcare dollars go. Under federal medical loss ratio (MLR) rules, insurers must spend at least 80% of premium dollars on patient care. Therefore, the bulk of healthcare spending occurs in clinical care: diagnosis, treatment and follow-up. Peer-reviewed studies show that large language models can match or exceed clinicians in performing many of these tasks. Yet these tools are rarely used for those purposes.
- Misaligned financial incentives. Under healthcare’s prevailing fee-for-service model, clinicians and hospitals are paid for each visit, test, procedure and hospital day. Thus, the more doctors and hospitals do, the more they are paid, regardless of whether the intervention adds value. A system structured this way has little financial incentive to eliminate unnecessary services or reduce utilization. The result is higher costs without improvements in clinical outcomes.
- Outdated care-delivery models. Unlike most industries, medicine remains anchored in old routines. For example, many hospitals postpone non-urgent tests and procedures over weekends. As a result, patients admitted on Friday on average stay an extra day compared to those admitted earlier in the week. That additional time increases costs without improving outcomes.
Conclusion 2: Healthcare will become even less affordable
In January 2026, the U.S. added 130,000 jobs, with roughly 95% of net new positions coming from healthcare and related roles. February’s report (released in early March) slightly tempered that trend due to a temporary nurses’ strike in California. Overall, healthcare continues to create most of America’s new jobs.
Now, if expanding the healthcare workforce actually made Americans healthier, we would see measurable improvements in long-term outcomes: fewer preventable complications from chronic diseases, reduced hospital admissions and increased longevity.
Instead, we’ve seen life expectancy largely plateau since the start of this century while U.S. healthcare expenditures tripled (from $1.4 trillion in 2000 to more than $5.3 trillion today).
Each year, medical care becomes less affordable for:
- Businesses, which insure roughly half the U.S. population. The average cost of covering a worker’s family now exceeds $25,000 per year, forcing companies to shift more of the burden to employees through salary freezes and higher deductibles.
- Families, who increasingly delay or forgo care because of cost. Medical debt remains a leading cause of bankruptcy.
- Government programs like Medicare and Medicaid, which consume growing portions of federal and state budgets, crowding out other public priorities.
When medical costs rise faster than payers can afford, employers and governments respond by narrowing coverage. Reduced access to preventive care and delayed treatment lead to more advanced illness and higher long-term expenditures. This vicious cycle sets the stage for the third conclusion.
Conclusion 3: Healthcare is ripe for disruption
In The Innovator’s Dilemma, Harvard Business School professor Clayton Christensen described how inefficient industries are upended as new technology become available. Generative AI is poised to do this in healthcare by simultaneously improving quality and lowering costs.
Consider three opportunities:
Chronic disease monitoring
Generative AI systems can analyze data from wearable monitors for patients with hypertension, diabetes and heart failure, identifying when conditions are poorly controlled and recommending medication adjustments. Better control of chronic disease could prevent more than 30% of heart attacks, strokes and kidney failure, according to CDC estimates, saving tens of thousands of lives and hundreds of millions of dollars a year.
Hospital-at-home models
Many short hospital stays could be replaced with home-based care supported by continuous GenAI monitoring and centralized telemedicine oversight. This approach would allow high-quality care to be delivered 24/7 at lower cost and without the risk of a hospital-acquired infection.
Early detection of inpatient deterioration
AI tools can continuously analyze real-time data from bedside monitors to detect subtle signs of clinical decline. Earlier intervention would prevent complications, reduce the need for intensive care and save lives.
The application of generative AI in medicine will likely begin where clinician shortages are most severe (rural communities and under-resourced regions). But when technologically supported models demonstrate better outcomes at lower cost, adoption will spread rapidly.
Recent employment data indicate a healthcare system nearing a crisis point. Doctors and hospitals can continue to meet growing demand by hiring more people. Or they can embrace technology, redesign care delivery and improve productivity. But they won’t be able to stop disruption.
Every industry Christensen chronicled wished it had acted sooner. Time will tell whether American medicine does the same.
