The Congressional Budget Office (CBO) has run the math on Obamacare again and has calculated that, due to the health care reforms, the size of the American workforce will decrease by two million full-time jobs in 2017, when all provisions of the law are in place.

And that’s:

• Great news! Workers aren’t chained to a job as the only viable way to get health insurance.
• Terrible news! It encouraged people to stop working and depend on government subsidies.
• Ambiguous news! It probably depends a lot on whether workers are choosing prudently.

We live in a time when statistics are easy to come by, but their interpretation is far more complex. For one thing, the CBO didn’t find that two million people will quit their full-time jobs. They summed up the total person-hours by which we’ll cut back, so, it could just as well be the result of four million people cutting their hours from full-time to half-time. Or ten million people shifting to a four day workweek.

But even if we’d nailed down the effect more precisely, our course of action wouldn’t be clear. The CBO and other government agencies have picked up the statistical tools of epidemiologists and big data groups like Facebook, but the outcomes they’re tracking aren’t as clear cut as whether mumps cases are going up or down or which ad produces the most clickthroughs.

It’s not as immediately obvious whether the CBO calculations are good news or bad news. If we learned that, as a result of Obamacare, two million people needed to find work in order to manage their health care costs, or that four million needed to increase their hours, we might interpret that as a poor outcome. Or, if regulations were prompting employers to fire two million workers or slash their hours (as we worried when businesses realized that cutting workers hours exempted them from providing benefits) the loss of these workers might look like bad news. After all, then people’s desire to work would be frustrated.

But the judgment of the CBO is that the reduction in labor will be mostly voluntary as workers will be the ones choosing to resign or reduce hours; it won’t be bosses forcing layoffs. The mere fact that a shift is voluntary, though, is not enough to tell us if it is desirable. Soft paternalist policies (the mandate itself being a case in point) are meant to put a thumb on the scale when citizens make choices, although the actual choice of whether to buy health insurance or pay a fine is technically up to the individual.

When nudges are at work, people seldom connect their choice with the regulatory constraint that informed it, so CBO actuaries may know the causes of their choices better than they do. Thus, it’s harder to rely on feedback from voters about how a law is affecting their lives. So, how are we to judge the merits of the change?

It’s tempting to amass more numbers. If a decline of two million jobs isn’t enough to make a normative claim, we could track the impact of the voluntary exodus using psychological surveys to track the prevalence of depression, or we could do a longitudinal study to discover the impact these choices had on the educational attainment of the children of these families, or we could use our modeling tools in a manner closer to their original use and look at health outcomes.

But we’re really bottoming out at a moral claim, that can’t be supported by facts and figures alone.

Is access to health care more or less important than making sure people are secured by their own labor, instead of a subsidy? What benefits does employment offer, besides the ability to purchase necessities and luxuries?

Politicians should make their cases explicitly based on the values their policies serve, without hiding behind purely empirical organizations to hint at their argument for them.

Follow @leahlibresco