What would a business-cycle balanced budget rule look like in Illinois?

by Matt Mitchell on May 5, 2014

in Balanced Budget, Institutions

A few years ago, I testified before the U.S. House Judiciary Committee. I’d been invited to talk about the design of a federal balanced budget amendment and much of my testimony drew on the lessons offered from state experience. Since 49 of the 50 states have such requirements, and since these requirements vary from state to state, I noted that federal lawmakers could learn from the state laboratory.

The best requirement, I argued, would have the following characteristics:

  1. Require balance over some period longer than a year. This effectively disarms the strongest argument against a balanced budget amendment: namely, that it would force belt-tightening in the middle of a recession. In contrast, if budgets need to balance over a longer time period, then Congress is free to run deficits in particular years as long as they are countered by surpluses in others.
  2. Allow Congress some time to come into compliance. You don’t have to be a Keynesian to worry that a 45 percent reduction in the deficit overnight might be a shock to the system.
  3. Minimize the gamesmanship associated with revenue estimation: Across the country, states with balanced budget requirements have to estimate revenue throughout the year (I’m a member of Virginia’s Joint Advisory Board of Economists and our responsibility is to pass judgment on the validity of these estimates). But this invites all sorts of questions: what model to use for the economy, should revenue be scored dynamically or statically, etc. One way to sidestep all of these questions is to make the requirement retrospective: require that spending this year not exceed revenue from years past.

Michigan Republican Justin Amash has proposed an amendment along these lines. It would be phased-in over 9 years and from there on out would stipulate that outlays “not exceed the average annual revenue collected in the three prior years, adjusting in proportion to changes in population and inflation.” Because it requires balance over three years rather than one, Amash calls it the “business cycle balanced budget amendment.”

Writing in Time, GMU’s Alex Tabarrok points to Sweden’s positive experience with a similar rule. And economists Glenn Hubbard and Tim Kane also endorse such a rule in their book, Balance.

Now, some Illinois state lawmakers have put together a proposal for a state rule that appears to be largely based on this model. It requires:

Appropriations for a fiscal year shall not exceed the average annual revenue collected for the 3 prior years, adjusting in proportion to changes in population and inflation.

(Unlike the Amash plan, however, the Illinois plan is not phased in over a number of years. Rather, it takes effect immediately upon passage of the bill.)

To see how it might work in a state, I decided to take the Amash Amendment for a test drive, using Illinois data. The solid blue line in the figure below charts Illinois’s actual general revenue from 1990 to 2012 in billions of current dollars. The dashed blue line phases in an Amash-type “business cycle” balanced budget rule. Once fully phased-in, it would limit spending to the average revenue of the three previous years, with an adjustment for inflation and population growth.


Notice three things:

  1. From 1990 to 2002, and from 2004 to 2007, the rule would have kept Illinois spending in line with Illinois revenue, and would have even allowed the state to run surpluses.
  2. In lean years (like 2008) when revenue levels off, the limit actually continues to rise. That’s because it is based on a longer time trend. This means that it wouldn’t require the sort of draconian budget cuts that balanced budget critics often fear. The accumulated surpluses from previous years could also be used to soften the blow.
  3. Lastly, note the (9 percent) revenue uptick from 2011 to 2012. The amendment would prudently make legislators wait a few years before they can go out and spend that money.

Previous post:

Next post: