EVERY SO often, the federal government’s financial performance surprises on the upside. Recent case in point: new data in the annual report of the Social Security trustees showing unexpected resiliency in the Social Security Disability Insurance (SSDI) program.

Just three years ago, in July 2015, the Obama administration warned that the program’s reserves were so low that it might not be able to cover expected benefits in 2016. Now, the trustees say the program will be solvent until 2032. Declining disability insurance receipts may be one reason that labor force participation by “prime-age” workers, those between the ages of 25 and 54, has ticked up from 80.6 percent in September 2015 to 81.8 in May 2018.

No matter how you look at it, this is good news, and it’s important to understand what it means. The main factor is that the labor market is far more robust than anticipated, producing an abundance of jobs for the lower-skilled, lower-income, working-age adults who were most likely to seek disability benefits as an alternative to employment during recessions, especially the brutal downturn that began in 2008. At approximately 1.5 million per year, annual applications for SSDI have now dropped slightly below their pre-recession levels, and the total number of workers on disability has been shrinking slightly but steadily since hitting a peak of 8.96 million in 2014. Total payout under SSDI has held steady at roughly $143 billion per year for the past three years, meaning that it has been shrinking as a share of both total economic output and overall federal spending.

Other causes include movement of beneficiaries from SSDI to the old-age retirement component of Social Security; retraining of administrative judges who were approving anomalously high rates of applicants; and Medicaid expansion, which made health care more widely available without going on disability, according to the New York Times.

What does not explain the decline is any structural reform to the program. The fact that disability rolls decline when the economy improves, and vice versa, reflects no intended purpose of disability insurance, because there’s no intrinsic connection between macroeconomic conditions and the likelihood of becoming disabled. Instead, SSDI has functioned as de facto long-term unemployment insurance, fraught with inefficiencies and perverse incentives. In particular, SSDI’s rules require that applicants be unable to engage in any significant paid work, giving them every incentive to cease working completely to qualify and to avoid rehabilitation — that is, to exit the labor force for good. The rules need to change so applicants face something other than a binary choice between work and benefits, perhaps by allowing benefits to phase out gradually as earnings from employment rise.

Undoubtedly, there will be those who greet the good news about SSDI’s solvency and decide “if it ain’t broke, don’t fix it.” In fact, there’s never a bad time to improve any federal program, and a moment when it has safely exited a crisis might actually be the best.

Read more: