How Congress Turns Deficit Reduction Into a Videogame

| Tue Oct. 9, 2012 12:06 PM EDT

Matt Yglesias says today that fixing a short-term deficit problem is something Congress can do. They can raise taxes or cut spending, as they did in 1990 and 1993, and deficits will go away. But it's a different story for long-term deficits. "No matter how much we scrunch our eyes together and promise really really hard, we can't force the political system of 2040 to avoid overspending on health care servies for my eighty-something dad rather than on private sector capital investments that will increase the productivity and wages of a hypothetical thirty-something kid. It simply can't be done."

More on this in a bit, but Matt goes on to make the perceptive point that since legislating long-term deficit reduction is all but impossible, self-proclaimed deficit hawks instead simply play a different, and completely useless, game:

What we can tackle is the shallow problem of CBO scores. Right now, the way the CBO scores things shows gigantic future deficits. If you pass a law saying "if the cyclically-adjusted budget deficit goes above 3.5% of GDP, Medicare reimbursement rates will automatically fall to eliminate the deficit" then you solve the CBO score problem.

It goes away like magic. But that's a dumb plan. Or, rather, it's not a plan at all. It's just a scoring rule. But if you peer into the details of different deficit reduction plans this is how they all work....The CBO doesn't employ fortune tellers who can assess conjectures about the future application of information technology to health care, about the military situation in the Pacific Rim, or about the political economy of tweaks in program design. So all the long-term plans end up relying on scoring rules. You direct the CBO to assess a situation in which congress "isn't allowed" to spend more than X on domestic programs or Y on the military or automatically applies cuts to hospitals. But giving the CBO those instructions doesn't change anything in the world, it's just an accounting exercise.

This is an extremely worthwhile point to internalize. A vast amount of what passes for long-term deficit reduction on Capitol Hill is just handwaving. This is especially true of Paul Ryan's endless parade of deficit reduction plans, all of which rely on telling CBO to assume various spending caps but provide no details on how those caps might be met or how they can possibly be enforced. When you do the arithmetic on these plans you always end up with wildly absurd projections, but that doesn't matter. CBO has to assume they're true, and the result is a series of nice charts showing the deficit slowly going away.

That said, I want to dissent slightly from Matt's point. There are, roughly speaking, four broad areas where Congress can act:

  • Taxes. Tax cuts and increases can't be forced on future generations and can't be prohibited either. As Republicans demonstrated in 2001, tax changes can be successfully passed via reconciliation, which means a simple majority in Congress can change tax law anytime it wants.
  • Discretionary spending. Ditto.
  • Social Security. This is a different case. Past experience suggests that when Congress passes rules that affect either future benefits or the future trajectory of payroll taxes, they stay intact. Practically speaking, Congress really can bind future generations here.
  • Medicare. This is a mixed bag. Some reforms don't stick (like the doc fix, which is deferred every year), while others do. Congress does have a limited ability to bind future generations to Medicare reforms that it passes today. However, the key problem here isn't really Medicare per se anyway. It's healthcare, full stop. Until we get a handle on that, there's no real chance of reining in Medicare growth.

Medium-term tax and spending changes aren't binding, but they do set the stage for future Congresses. Put that alongside long-term changes to Social Security and Medicare, and in practical terms you really do have a limited amount of power to bind future Congresses. It's not perfect by any stretch, but it's not entirely useless either.