The start of this decade saw a remarkable change in Americans’ health: statistically speaking, it began to weaken.

To be more accurate, the general improvement in Americans’ life expectancy at birth, which dates back at least to 1960 and has been fairly steady since 1980,[1] lost momentum in this decade.[2] And in 2015, the most recent year for which data are available, American life expectancy actually fell.[3] The decrease was small—one-tenth of a year—but such declines are very rare; the last one occurred in 1993.[4] The decrease follows four straight years of some of the weakest improvements in life expectancy of the last four decades.[5]

More worrisome, the data indicate the decline isn’t the result of a spike in one or two causes of death. Mortality is up nearly across the board: heart disease, respiratory disease, accidents, stroke, Alzheimer’s, diabetes, and suicide.[6] Demographically, it is affecting white males, white females and black males, and is concentrated among young and middle-aged Americans. And it is U.S.-focused; other developed nations are experiencing significant advances in life expectancy.[7]

It’s tempting to wonder if this sad trend has something to do with another major change involving Americans’ health over the same time: the implementation of the Patient Protection and Affordable Care Act (ACA), more commonly called “Obamacare.” The first provisions of the ACA went into effect in 2010 and most major provisions were implemented by January 2014, though some final pieces are to be phased in by 2020.

There currently are too little data to test for a correlation between the ACA and weakening U.S. life expectancy. It’s entirely possible the past few years’ slump will prove to be statistically unconnected to the ACA. Indeed, some worrisome trends underlying U.S. life expectancy date back two decades.

But the slump underscores something that few people—including ACA supporters—realize: there’s surprisingly little correlation between expanding health insurance coverage—including some types of ACA-mandated preventive care and screenings[8]—and improved health.[9]

The ACA has significantly reduced the number of people without coverage. In 2010, more than 47 million Americans lacked coverage for at least part of the year, while less than 30 million were similarly uncovered in 2015.[10] But if expanded coverage means a healthier population, then shouldn’t longevity now be surging instead of slumping?

Among health care experts, it’s commonly believed that about one-third of the money Americans spend on care yields no health benefits.[11] Moreover, Americans in need of care regularly find ways to get it, even if they don’t have coverage.[12] So it shouldn’t be too surprising if the ACA’s coverage expansion wouldn’t deliver much improvement in Americans’ health.

But if the ACA isn’t improving Americans’ health (and that’s still an “if” until we get more data), then what’s the virtue of the ACA?

The Trump administration and congressional Republicans have vowed to repeal the ACA and replace it with something. But they can’t agree on what that something is, as evidenced by last week’s aborted vote on the American Health Care Act that was drafted by GOP leaders. Some Republicans and many political commentators say that whatever is ultimately adopted should result in coverage rates at least as good as the ACA. Like ACA supporters, they assume that greater coverage means better health.

American health care does need major improvement—just as it did in 2010. Back then, people complained that health care was too expensive, constraining, complex and confusing, and seemingly filled with bad incentives that resulted in poor outcomes.

Problem is, there is no simple “fix” for health care. Most policy “solutions”—including the ACA and the recently aborted GOP proposal, as well as more radical proposals like single-payer and socialized medicine—impose on all Americans a specific set of good and bad tradeoffs decided by politicians and bureaucrats. Those tradeoffs concern such things as what sorts of care should be required and what should be optional or discouraged, what sorts of obstacles to care would be acceptable and what would not, and who should pay what costs.

In writing the ACA, the Obama administration and congressional Democrats—with the best of intentions—made one set of tradeoff decisions. The result has been a lot of people complaining that care is now too expensive, constraining, complex and confusing, and seemingly filled with bad incentives—that is, the same complaints that were made in 2010. If Republicans likewise adopt a system that simply embodies a different set of tradeoffs, similar complaints will be made about the GOP system.

To truly reform health care, lawmakers need to provide Americans with more options on care and coverage, more transparency on how health care works and greater respect for the many different demands on medicine that exist in an incredibly diverse nation of 325 million people. Lawmakers shouldn’t act in accordance with some artificial, politicized deadline to put a new plan in place. And they shouldn’t make far-fetched promises about the wondrous results their proposals will yield.

Health care policy is hard, as President Trump recently discovered.[13] For that reason, it’s important to leave many of the decisions concerning care in the hands of the individuals who will live under those decisions. The only promise on reform that Republicans—and Democrats—should make and keep is that they will nurture a health care marketplace that better responds to the many demands and choices of its consumers.