Google+
Close

The Agenda

NRO’s domestic-policy blog, by Reihan Salam.

How to Make Decentralization Work



Text  



I was struck by the similarities between Alberto Alesina’s prescriptions regarding the principles that should guide decentralization, which are rooted in Italy’s disastrous experiment with decentralization, and Michael Greve’s emphasis on the importance of “one problem, one sovereign” in competitive federalism. Having identified the theoretical benefits of decentralization — it can bring government closer to the people; it can lead to a more efficient allocation of responsibilities, with functions that have large economies of scale taken on by central governments and those that have small economies of scale taken on by local or regional governments — Alesina identifies two rules for keeping excesses in check:

First, localities should be required to run balanced budgets. Otherwise, their eligibility for bailouts from the national government will discourage them from spending within their means. Even with a balanced-budget rule, a regional government can be profligate if the national government decides to transfer money directly to it from another region (in addition to the indirect interregional transfers that necessarily result from progressive income taxes and national welfare programs). Any direct transfers of this kind should be made as transparently as possible. But the approach nevertheless tends to encourage waste and corruption—an exchange of favors between national politicians and local lobbies. A better approach is to finance regional spending with local taxes.

The second critical element of successful decentralization is a clear division of responsibility. In practice, this has often proved difficult to achieve. Tasks are allocated not according to efficiency but according to the result of battles among different levels of government. Duplication, confusion, and conflict among these levels of government lead to inefficiency, bureaucracy, and a rising tax burden. [Emphasis added]

In a similar vein, Greve writes:

Washington has indeed assumed far too much power over local affairs. In many policy areas, decentralization would improve our policies and our politics. I also agree that our problems have a potent constitutional dimension. And yet: I will try to persuade you that our federalism problem is a great deal more complicated. “Federalism” in the sense of decentralization and state authority is not just a solution. In many respects, it is also a problem and a cause of widely lamented institutional dysfunctions. Put differently: we actually have not too little but far too much federalism of the wrong kind—a federalism that makes government bigger, more irresponsible, and less accountable. We need less of that federalism, and more federalism of the right kind—a federalism that disciplines government at all levels and makes it more accountable and transparent. We will over the coming years re-negotiate our federalism bargain. But the central question is not, how much federalism? It is, what kind of federalism? 

And the problems with “our federalism,” according to Greve, are rooted in our departures from competitive federalism:

Matters that are genuinely national should be committed entirely to the federal government. National defense, the regulation of network industries (such as airlines or the internet), and perhaps large-scale redistribution (such as Social Security) are the classic examples. The assignment should be exclusive, so as to avoid the regulatory duplication, gamesmanship, and intergovernmental collusion that characterize our federalism. On the other side, lots of things should be left exclusively to the states—exclusive of any federal regulation, and exclusive of any federal funds. This principle limits the federal government and, more importantly, forces state competition. Each state’s citizens will have to tax themsleves for whatever level of public services and redistribution they want. When a state offers a lousy mix of taxes and services, productive citizens and businesses move to a more hospitable jurisdiction. Over time, one hopes, states that don’t give citizens their money’s worth will learn their lesson. [Emphasis added]

That is, we should have a clear division of responsibility and we also need to transfers to regional governments, i.e., states:

Equally pervasive are fiscal federalism cartels: instead of letting states fund programs from own-source revenues, under competitive conditions, Congress collects taxes from citizens across the country and then sends them to state and local governments. Medicaid, education, welfare, and countless others operate on this principle. The programs usually require states to comply with federal funding conditions and to match the federal funds, but they are driven by fiscal incentives. No state would devote 25 percent of its budget to Medicaid if it had to tax its own citizens for the cost. But states are evidently willing to tax themselves for half the cost—because if they don’t, they leave federal dollars on the table. And so every state ends up with a level of spending that no state would adopt on its own.

All this is to say that great minds think alike. For more on federalism, I strongly recommend Greve’s recent essay for the Duke Journal of Constitutional Law & Public Policy, titled “Our Federalism Is Not Europe’s. It’s Becoming Argentina’s.

The depressing is that hardly anyone cares about federalism, and those who do don’t fully understand Alesina’s two rules. Most self-described Tenth Amendment fans emphasize “state’s rights” over the principle that states ought to compete. One of the ironies of the refusal of at least 26 states to establish state-based insurance exchanges under the Affordable Care Act is that, intentionally or otherwise, it sets up a scenario in which the federal government has exclusive authority over medical insurance. And that would actually be better than setting up the insurance exchanges as essentially a second Medicaid program, with the same “duplication, gamesmanship, and intergovernmental collusion” that defines the first one. 



Text  


Subscribe to National Review