The folly of using money to improve health care quality

If money alone could buy better health care quality, we’d have solved this years ago with a well-timed wire transfer and a polite note that says, “Please deliver fewer complications. Thx.” Instead, the U.S. spends a lot, argues a lot, measures a lotand still finds itself asking why “quality” can feel like a mirage that moves every time we walk toward it.

The problem isn’t that money doesn’t matter. It does. Staffing, training, safe buildings, modern equipment, data systems, patient transportation, interpretersnone of that is free. The folly is believing that moving money around (bonuses, penalties, reimbursement tweaks, “pay-for-performance”) automatically produces better outcomes, safer care, kinder experiences, and healthier communities. In real life, quality is a system property. And systems don’t respond to incentives like vending machines.

Why “just pay more” sounds logical (and why it isn’t)

Quality isn’t a “more in, better out” machine

In manufacturing, paying for higher-grade materials can improve the final product. In health care, “quality” is built from thousands of decisions across clinicians, teams, workflows, and settingsoften under time pressure and imperfect information. Throwing money at one part of the chain can leave other links weak. A hospital can nail its surgical checklist and still struggle if post-discharge support is scarce, if primary care is overloaded, or if patients can’t afford medications.

Spending more doesn’t reliably translate to better outcomes

A stubborn body of U.S. research suggests that higher spending and higher utilization are not consistently associated with better quality or better outcomes. Sometimes high-spending regions deliver more tests, more visits, and more procedures without delivering more of the evidence-based care that actually moves the needle for patients.

Incentives can’t fix what the system can’t do

Financial incentives assume the problem is motivation: “If we pay for quality, clinicians and hospitals will do quality.” But many quality gaps are capability gaps, not willpower gaps. You can’t bonus your way into having enough nurses. You can’t penalty your way into interoperable data. You can’t “incentivize” a functional home health workforce into existence by Tuesday.

How quality metrics get weird in the wild

Goodhart’s Law: when a measure becomes a target, it stops being a measure

Quality programs often rely on metrics: readmissions, complication rates, patient experience scores, diabetes control, blood pressure control, screening rates, and more. Metrics are essential. But once money is attached, behavior changesand not always in the direction patients would hope.

The classic failure mode is “teaching to the test.” If the metric rewards what’s easy to document, documentation rises. If it rewards what’s easiest to improve, effort shifts there. Meanwhile, unmeasured problemspain control, diagnostic accuracy, respectful communication, equity, continuitycan be crowded out.

Process measures improve faster than outcomes (sometimes for boring reasons)

Many pay-for-performance programs show modest improvements in process measures (the “did we do the thing?” checkbox). Outcomes are harder. They are influenced by social factors, follow-up access, and patient circumstances. They also take longer to change and are noisierespecially for smaller hospitals or small physician practices.

Risk adjustment and fairness are not footnotesthey’re the whole story

Two hospitals can deliver equally skillful care and still look different on paper if one serves a population with higher medical complexity, fewer social supports, language barriers, housing insecurity, and limited access to post-acute services. If incentives don’t account for this, “paying for quality” can unintentionally become “paying for patient mix.”

Administrative burden is a quality problem, too

Incentive programs often require reporting. Reporting requires staff time. Staff time costs money. And staff time diverted to chasing metrics is staff time not spent on patient care, improvement work, clinician coaching, or redesigning broken workflows. The irony is painful: a program meant to improve quality can consume the very resources needed to improve quality.

Three cautionary tales from Medicare’s incentive era

The U.S. has been running large-scale experiments in financial incentives for years. Medicare alone has multiple programs that adjust payments based on performance. These efforts have taught us something important: incentives can create movement, but movement is not the same thing as improvement.

1) Hospital Readmissions Reduction Program (HRRP): fewer readmissions, complicated trade-offs

HRRP links hospital payment to readmission rates for certain conditions, with penalties capped at a fixed percentage. Hospitals responded. Many invested in discharge planning, follow-up calls, medication reconciliation, and better handoffs.

But HRRP also illustrates why “money = quality” is too simplistic. Readmissions are influenced by what happens after discharge: primary care access, home health reliability, transportation, caregiver support, medication affordability, and the availability of skilled nursing facilities. When a metric is tied to hospital payment but depends heavily on community capacity, the hospital ends up wearing responsibility it can’t fully control.

Even more importantly, researchers have raised concerns about unintended consequences, including whether pressure to reduce readmissions could lead to changes in care patterns that don’t benefit all patients equally. Some analyses have debated possible associations with adverse outcomes in certain groups, reminding us that incentives can shift behavior in ways that aren’t captured by the headline metric.

2) Hospital Value-Based Purchasing (HVBP): money shuffles, results don’t always follow

HVBP is designed as a budget-neutral program: money is withheld from hospitals and then redistributed based on performance across quality and cost domains. In theory, that’s elegantno new spending, just smarter spending.

In practice, HVBP reveals how hard it is to create a single score that truly reflects “quality.” Hospitals are complex organizations with competing priorities, and a multi-domain scoring system can create perverse incentives: improve the easiest domain, manage the scoreboard, and hope the total score works out. Meanwhile, evidence on whether such programs consistently improve patient experience and outcomesespecially across different hospital typeshas been mixed.

3) MIPS: when quality becomes a paperwork tournament

The Merit-based Incentive Payment System (MIPS) adjusts clinician payments based on performance across quality, cost, improvement activities, and electronic reporting. It aims to move practice toward value. That’s the mission statement.

The lived reality for many clinicians is that MIPS feels like a regulatory obstacle course where the prize is small, the rules change, and the main strategy is to avoid penalties. When quality measurement becomes a tournamentwinners and losers based on shifting thresholdsorganizations rationally invest in scoring tactics: measure selection, reporting workflows, registry participation, and compliance staffing.

Here’s the paradox: the more complex the incentive design, the more it rewards the ability to navigate complexity. Large, well-resourced systems can buy the infrastructure to play the game. Smaller practices can get squeezed. That’s not quality improvement; that’s quality accounting.

So what actually improves health care quality?

If using money as a lever can be foolish, what’s the smarter approach? The answer is not “never use money.” It’s “use money to build capability, not just to rearrange motivation.”

Invest in the fundamentals that make quality possible

• Staffing and retention: safe nurse staffing, stable teams, clinician well-being, and training. Burned-out staff can’t deliver consistently excellent careno matter how spicy the bonus.
• Care coordination capacity: transitional care nurses, pharmacists, social workers, community health workers, and warm handoffs that don’t rely on heroics.
• Data that clinicians can actually use: interoperable records, real-time dashboards built for clinical workflows, and feedback loops that help teams learn quickly.
• Primary care and prevention: the boring stuff that prevents emergenciesaccess, continuity, chronic disease management, vaccinations, and screenings done well.
• Community supports: transportation, food access, medication access, and partnerships with post-acute providersbecause “quality” doesn’t stop at the hospital doors.

Use improvement science, not just incentive math

High-performing health systems behave like learning organizations: they identify problems, test changes, measure results, scale what works, and create a culture where staff can speak up about risks. That requires leadership, time, and psychological safetynot just a financial formula.

Design incentives that support learning (and avoid predictable traps)

If you’re going to attach dollars to quality metrics, treat it like a clinical intervention: start with a theory of change, anticipate side effects, and monitor what happens. Some practical guardrails:

• Measure fewer things, better: focus on outcomes that matter, avoid metric sprawl, and retire measures that are “topped out” or easily gamed.
• Reward improvement, not just rank: don’t punish hospitals for serving hard-hit communities; recognize progress from a baseline.
• Build in equity protections: stratify results, watch disparities, and avoid designs that financially kneecap safety-net providers.
• Keep incentives large enough to matterbut not so large they distort care: the point is to nudge attention, not to create panic-driven medicine.
• Pair incentives with resources: penalties without support can become a slow-motion “rich get richer” cycle in health care infrastructure.

Bottom line: money is necessary, but it’s a terrible substitute for a working system

The folly isn’t spending money on quality. The folly is treating money like a magic spell: “I hereby declare quality improved!”and expecting the system to comply. Quality improves when people have the tools, time, training, data, and teamwork to deliver reliable care for every patient, every day.

Financial incentives can play a role, but they’re best used as a supporting actor, not the lead. Otherwise, we end up with a health care system that gets very good at measuring itselfand only occasionally gets better at caring for people.

Experiences related to “The folly of using money to improve health care quality” (extra)

Ask almost anyone who’s worked inside an incentive-driven quality programclinicians, nurses, administrators, quality analystsand you’ll hear a familiar pattern: the intent is admirable, the experience can be… complicated.

The clinic that “improved quality” by improving its click-speed

In many primary care practices, the first visible change after a new pay-for-performance push isn’t suddenly perfect diabetes control. It’s workflow redesign around documentation. Teams build templates. Medical assistants get training on which boxes must be checked. Someone creates a spreadsheet of patients who are “close to target” so the clinic can focus outreach on those most likely to move the metric. Quality improveson paperbecause the practice got better at capturing what it already did. That’s not nothing. But it’s also not the same as fewer strokes, fewer heart attacks, or fewer amputations.

The nurse who became a part-time detective

In hospitals responding to readmission penalties, nurses often describe the whiplash of being asked to do more care coordination without more time. Discharge planning becomes an investigative project: Who’s picking the patient up? Do they have stairs at home? Can they read the instructions? Is the pharmacy open? Do they have a scale to monitor heart failure weight changes? Incentives can elevate the importance of these questions (good!), but when resources don’t follow, the workload becomes a daily scramble powered by sticky notes and heroism. Eventually, the system starts depending on the heroism. That’s where quality initiatives quietly turn into burnout machines.

The hospital that learned “don’t be bad at measured things”

Some hospitals describe an awkward evolution: they don’t become “excellent,” they become “less penalizable.” They add a unit to improve one metric while another metric drifts. They prioritize conditions tied to penalties and pay less attention to equally serious problems that aren’t on the scoreboard this year. You can feel the culture shift when dashboards dominate meetings and staff sense that the metric matters more than the patient story behind the metric. Not because leaders are villains, but because incentive programs can make the score feel like the mission.

The safety-net reality: being punished for serving the hardest cases

In organizations serving patients with fewer social supports, quality leaders often talk about the frustration of trying to improve outcomes that depend on the local ecosystem. If housing is unstable, medication adherence is harder. If transportation is unreliable, follow-up appointments get missed. If community mental health resources are scarce, crises escalate. When payment penalties hit these hospitals, it can reduce the very capacity needed for improvementfewer care managers, fewer community partnerships, fewer transition-of-care calls. That’s a vicious cycle: the hospitals with the most improvement work to do can end up with fewer resources to do it.

When money helps: funding capability instead of just rewarding results

The most encouraging stories tend to sound less like “we dangled a bonus” and more like “we built a better system.” For example: a program funds pharmacists to run medication reconciliation calls within 48 hours of discharge; a health system invests in bilingual care navigators; a payer supports home-based follow-up for high-risk patients; a clinic gets protected time for team huddles and rapid-cycle improvement. In these cases, money doesn’t “buy” quality directlyit buys the conditions in which quality can reliably emerge.

That’s the lived lesson behind the folly: incentives can shift attention, but attention isn’t the same as capacity. If we want durable health care quality improvement, we should spend less energy trying to bribe the system into being betterand more energy building a system that can be better.

Conclusion

Using money to improve health care quality isn’t inherently wrongit’s just frequently oversold. Financial incentives work best when they’re simple, fair, and paired with real operational support. Without that, they can drive paperwork, gaming, inequity, and a narrow focus on whatever happens to be measured. The smarter path is to invest in people, workflows, data, and community supports, then use measurement to learnnot just to pay.