Institutions and state spending: an overview.

• Author: Mitchell, Matthew
• Position: Essay

U.S. fiscal policy at the federal, state, and local level is on an unsustainable path. (1) Although reformers look for ways to reduce spending on particular budget items, tomorrow's legislatures may easily reverse these cuts. In contrast, a change in the rules that govern the political process--the "institutions" that shape a budget--can have a lasting effect on spending for years to come. (2) Codified in statutes and in constitutions, these institutions include the rules of budgeting, electioneering, and legislating. They influence the decisions of legislators, governors, presidents, bureaucrats, voters, and even lobbyists. Institutional reform can be a more effective and sustainable path to fiscal probity than a one-time budget cut. In this article, we summarize the empirical investigations of more than a dozen state-level institutions. The lesson for both state and federal policymakers is that a number of institutional reforms are more likely to put spending on a more sustainable path.

The Laboratory

Justice Louis Brandeis famously referred to the federal system as a "laboratory" in which each state is free to implement novel social and economic experiments (New State Ice Co. v. Liebmann, 285 U.S. 262 [1932]). For the social scientist interested in understanding how institutions affect policy outcomes, the metaphor is apt. Each of the fifty states has its own set of institutions, some of which have changed over the past several decades. At the same time, many other factors that might influence outcomes do not vary across the states. In other words, cross-state studies effectively control for factors such as macroeconomic conditions, culture, and the broad legal/ constitutional setting in which each state operates. Moreover, researchers can employ various econometric techniques to control for the influence of factors that do differ across states, such as climate and demographic makeup. In sum, the state setting provides a rich laboratory in which to test the effect of different institutions on spending.

Many researchers have performed such tests. Here we review this literature, summarizing the effect of sixteen institutions on state spending. (3) We concentrate on peer-reviewed studies that rely on large data sets and well-accepted econometric techniques. Each of these studies carefully accounts for other factors that have been shown to affect spending, such as demographic makeup, climate, and the economy.

We begin with some of the most-studied institutions (strict balanced-budget requirements), move on to those that have received comparatively less attention (for example, item-reduction vetoes), and conclude with a pair of institutions whose study has yielded conflicting conclusions (direct democracy and term limits). Where possible, we present the findings in terms of the institutions' effect on spending per capita and convert all dollar amounts to 2008 dollars to permit side-by-side comparison. We hope that this overview will provide useful insights to state policymakers who may be interested in reform. We also hope that it will be helpful to federal policymakers because many of these institutions may be transferred to the federal level.

Figure 1 highlights the results of the studies that present their findings in terms of spending per capita. To put some of these figures in perspective, state and local expenditures per capita in 2008 averaged approximately $5,708. (4) Thus, the studies we review suggest that these institutions can decrease spending per capita from 1 to 22 percent. Justice Brandeis's "experiment" clearly has yielded economically significant results.

Sixteen Institutions and Their Effects on State Spending

1. Strict Balanced-Budget Requirements

   With the exception of Vermont, every state has a balanced-budget requirement, but the stringency of these requirements varies widely. For example, in some states the governor must submit a balanced budget, but the legislature need not pass one. In other states, estimates of the enacted budget need to show balance, but there are no consequences if these estimates prove wrong at the end of the fiscal year and the actual budget is not in balance. In some states, the legislature may carry over a deficit from one year to the next, but in others they may not. Finally, in some states an independently elected Supreme Court is the ultimate enforcer of the requirement, whereas in others the legislature appoints the members of the Supreme Court. Even though balanced-budget requirements are designed to ensure only that spending be less than revenue, these requirements may reduce both spending and revenue. James Buchanan and Richard Wagner (1977) explain why both might be reduced by observing that the ability to buy items for current taxpayers while leaving future taxpayers to pick up the tab systematically biases policy in favor of greater spending.

   [FIGURE 1 OMITTED]

   Indeed, several studies have found that states with stricter balanced-budget requirements tend to tax and spend less than other states. Henning Bohn and Robert Inman (1996) for example, find that spending per capita is approximately $189 less in states with strict balanced-budget requirements relative to those with only weak requirements. (5) David Primo (2007) arrives at a remarkably similar result, finding that strict balanced-budget requirements reduce spending per capita by approximately $184 per capita. (6) If this amount is the effect of moving from a weak to a strict balanced-budget requirement, it is possible that if the federal government were to adopt a balanced-budget requirement, the impact would be even greater.

   Strict balanced-budget requirements have other salubrious effects. Bohn and Inman (1996) also find that states with strict requirements tend to have larger rainy-day funds and larger surpluses and that states with these requirements tend to balance their books through spending reductions rather than with revenue increases. Shanna Rose (2006) finds that states without strict balanced-budget requirements are more likely to suffer from a "political business cycle" whereby policymakers increase spending shortly before an election, only to cut back after the election.

   A strict balanced-budget requirement may have some unintended consequences. Noel Johnson, Matthew Mitchell, and Steven Yamarik (2011) conclude that although such rules limit the likelihood of partisan fiscal outcomes, they increase the likelihood of partisan regulatory outcomes. When Democratic-controlled states were unable to carry a deficit forward to the next fiscal cycle, they were more likely to raise the minimum wage, less likely to adopt a right-to-work statute, and more likely to regulate personal freedoms.

2. State Rainy-Day Funds

   Some commentators worry that a balanced-budget requirement exacerbates the ups and downs of the business cycle (see, for example, Ornstein 2011). Because state budgets tend to be tightest at the bottom of an economic downturn, this argument goes, strict balanced-budget requirements force states to cut back on spending at the worst time. One institutional answer to this critique is a "budget-stabilization fund," better known as a "rainy-day fund." States contribute to this fund during good years and draw on it when the budget is strained owing to a downturn or some other event, such as a natural disaster. Forty-seven states currently maintain such funds, but, as for many institutions, their design varies from state to state (Rueben and Rosenberg 2009).

   Studies of rainy-day funds suggest that they can smooth out the spending cycle, but the details matter. Gary Wagner and Erick Elder's (2005) study is the most comprehensive recent examination of rainy-day funds. They find that states whose rainy-day funds have strict rules governing deposits and withdrawals tend to experience less volatility of spending per capita ($14 less, as measured by the cyclical variability of spending per capita over time).

3. Supermajority Requirements for Tax Increases

   Fifteen states require supermajority votes of the state legislature in order to raise taxes (Waisanen 2010). Depending on the state, these rules require two-thirds, three-quarters, or three-fifths of the legislature to consent to a tax increase. In some states, the supermajority requirement applies to all taxes, and in others it applies to a...