Continuing our series on budgetary reversions and government shutdowns, I am pleased to introduce David Dreyer Lassen, a professor in the Department of Economics at the University of Copenhagen.


Recent posts on Monkey Cage have examined why (and where) you’d ever have a political system that allows for late budgets and costly government shutdowns.

Outside of the U.S. federal government, Australia has been mentioned, but of course you need only to look to state capitols to see scores of late budgets, government shutdowns – and creative solutions to try to mitigate the costs of such shutdowns.

In the graph below, adapted from data used in Andersen, Lassen and Nielsen 2012, we show the distribution of the number of days from the end of the fiscal year to final budget enactment for most state governments 1988-2007. A negative number means the budget was on time, a positive number that it was late. In our sample, almost a quarter of budgets were late, and many were seriously late.

Graph by David Dreyer Lassen
Graph by David Dreyer Lassen

The main reason for late budgets, unsurprisingly, is the presence of divided government, where the budget becomes a bargaining chip between an executive and a legislature controlled by different parties (the point that divided government is important for late budgets was made simultaneously by Klarner et al. (JoP, 2012) on a different sample).

In his post on the rationale behind a system that allows government shutdowns as threat points in political bargaining, Gary Cox notes that – absent executive control over the budget or a transition to parliamentarism – an alternative to the current system would be to lower the costs that arise when budgets are not passed on time.

State governments differ considerably with respect to what happens if a budget is late. Some states have no idea, since they have never experienced it (a condition not unlike breaching the debt ceiling); one state budget director told us that his state had never had – nor would ever have – a late budget. In some states, government shuts down, and if there are no money on the books, they resort to IOUs. Many states that have experience, sometimes substantial, with late budgets are indeed the states with provisions that ease the consequences of late budges. Some of these states have temporary budgets that automatically kick in if no budget is in place, while others have legislated – or occasionally court ordered – continuing payments for select functions, including schools, roads and, in one case, psychiatric hospitals and schools for the deaf and the blind.

What are the consequences of such attempts to lower the costs of bargaining in a presidential system? States that have such provisions experience both a higher frequency of late budgets and longer delays, conditional on the budget being late in the first place. For a given benefit of winning, lower costs of an additional delay makes it worthwhile to wait longer to see if one’s opponent gives in – and the end result is longer delays. This kind of logic would be consistent with Republicans’ current attempts to fund select popular or important programs, which, if they were successful, would only serve to drag out the negotiations even more.

However, this may not be the whole story. We observe that states that shut down absent a budget almost never have shutdowns. This could of course be because shutdowns are costly or, and we actually think this is more likely, it could be because states that for some other reason (polarization, as suggested by results in another draft paper of ours, or other budget process institutions) are late eventually pass laws that make late budgets less costly and – thus – that the causality actually runs from late budgets to institutional changes.

Oh, and our results show that budget passage is much swifter in election years, which doesn’t bode well for the current situation.