Editor’s note: This is an excerpt from a new book, American Federalism Today: Perspectives on Political and Economic Governance, edited by Michael J. Boskin and available from the Hoover Institution Press.
Delegates to the Constitutional Convention in Philadelphia in the summer of 1787 eventually emerged with a structure described by James Madison as “partly national, partly federal.” This contemplated a genuinely national government, with representation from the people (and not just the states) and power to enforce its own laws through a vigorous executive and an independent judiciary, but the states would retain political autonomy and authority over the issues most significant to ordinary life. The powers of this national government would be confined to certain enumerated objects, primarily foreign affairs and interstate commerce. This was an innovation; there were no precedents in world history for such a mixed system.
The “natural attachment” of the people in 1787 to their states, as Madison described it in Federalist no. 46, was powerful—far more so than today. But the framers of the federalist system were not content to rest on natural attachments alone. They offered practical and theoretical arguments about how the new system of dual sovereignty would promote three complementary objectives: (1) “to secure the public good,” (2) to protect “private rights,” and (3) “to preserve the spirit and form of popular government.” Achievement of these ends, according to James Madison, was the “great object” of the Constitution. To understand the founders’ design, we must look again at those arguments—not just in the mouths of the Federalists, who prevailed, but of the Anti-Federalists, too. As the people of the twenty-first century, we must evaluate these arguments in light of modern experience and knowledge about political decision making. Many of the arguments of 1787 stand up remarkably well, but others do not.
To “secure the public good”
Rejecting both pure confederation and consolidation, the “Federal Farmer” (a particularly able and influential Anti-Federalist pamphleteer) argued that a “partial consolidation” is the only system “that can secure the freedom and happiness of this people.” He reasoned that “one government and general legislation alone, never can extend equal benefits to all parts of the United States: different laws, customs, and opinions exist in the different states, which by a uniform system of laws would be unreasonably invaded.” Three important advantages of decentralized decision making emerge from an examination of the founders’ arguments and the modern literature. First, decentralized decision making is better able to reflect the diversity of interests and preferences of individuals in different parts of the nation. Second, allocation of decision-making authority to a level of government no larger than necessary will prevent mutually disadvantageous attempts by communities to take advantage of their neighbors. And third, decentralization allows for innovation and competition in government.
One size does not fit all. So long as preferences for government policies are unevenly distributed among the various localities, more people can be satisfied by decentralized decision making than by a single national authority. This was well understood by the founding generation. States are preferable governing units to the federal government, and local government to states. Modern public-choice theory provides strong support for the framers’ insight on this point.
A second consideration in designing a federal structure is more equivocal. The unit of decision making must be large enough so that decisions reflect the full costs and benefits, but small enough that destructive competition for the benefits of central government action is minimized. In economic language, this is the problem of externalities.
Externalities present the principal argument for centralized government: If the costs of government action are borne by the citizens of state C, but the benefits are shared by the citizens of states D, E, and F, state C will be unwilling to expend the level of resources commensurate with the full social benefit of the action. This was the argument in Federalist no. 25 for national control of defense. Because a Minuteman III missile in Pennsylvania will deter a Russian or Chinese attack on Connecticut and North Carolina as well as Pennsylvania, optimal levels of investment in Minutemans require national decisions and national taxes. Similarly, because expenditures on water pollution reduction in Kentucky will benefit riparian zones all the way to New Orleans, it makes sense to regionalize or nationalize decisions about water-pollution regulation and treatment. Thus, as James Wilson explained to the Pennsylvania ratifying convention, “Whatever the object of government extends, in its operation, beyond the bounds of a particular state, should be considered as belonging to the government of the United States.”
That significant external effects of this sort provide justification for national decisions is well understood—hence federal funding of defense, interstate highways, national parks, and medical research; and federal regulation of interstate commerce, pollution, and national labor markets. It is less well understood that nationalizing decisions where the impact is predominantly local has an opposite effect. If states can obtain federal funding for projects of predominantly local benefit, they will not care if total cost exceeds total benefit; the cost is borne by others. The result is a “tragedy of the commons” for Treasury funds.
The framers’ awareness that ill consequences flow as much from excessive as from insufficient centralization is fundamental to their insistence on enumerating and thus limiting the powers of the federal government. Hence the other half of Wilson’s explanation: “Whatever object of government is confined in its operation and effect, within the bounds of a particular State, should be considered as belonging to the government of that State.” This stands in marked contrast to the modern tendency to resolve doubts in favor of federal control.
A final reason why federalism may advance the public good is that state and local governmental units will have greater opportunity and incentive to pioneer useful changes. Justice Louis Brandeis put the point most famously: “It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.” A consolidated national government has all the drawbacks of a monopoly: it stifles choice and lacks the goad of competition. If innovation is desirable, it follows that decentralization is desirable.
Perhaps more important is that smaller units of government have an incentive, beyond the mere political process, to adopt popular policies. If a community can attract additional taxpayers, each citizen’s share of the overhead costs of government is proportionately reduced. Since people are better able to move among states or communities than to emigrate from the United States, competition among governments for taxpayers will be far stronger at the state and local than at the federal level. Since most people are taxpayers, this means that there is a powerful incentive for decentralized governments to make things better for most people. In particular, the desire to attract taxpayers and jobs will promote policies of economic growth and expansion.
To be sure, the results of competition among states and localities will not always be salutary. The most important example of this phenomenon is the effect of state-by-state competition on welfare and other redistributive policies. In most cases, immigration of investment and of middle- to upper-income persons is perceived as desirable, while immigration of persons dependent on public assistance is viewed as a drain on a community’s finances. Yet generous welfare benefits paid by higher taxes will lead the rich to leave and the poor to come. This creates an incentive, other things being equal, against redistributive policies.
This is an instance of the free-rider problem: even if every member of the community would be willing to vote for higher welfare benefits, it would be in the interest of each to leave the burden of paying for the program to others. Presumably that is why advocates of a more generous social safety net tend to push for expansion of federal programs, while advocates of the opposite policy tend to favor state-oriented solutions.
Thus, the competition among states has an uncertain effect: often salutary but sometimes destructive. There are races to the bottom as well as races to the top. And it is often impossible to know which is which; this will depend on substantive policy preferences.
To protect “private rights”
At the time of the founding, defenders of state sovereignty most commonly stressed a second argument: that state and local governments are better protectors of liberty. The most eloquent of the opponents of the Constitution, Patrick Henry, declared that in the “alarming transition, from a Confederacy to a consolidated Government,” the “rights and privileges” of Americans were “endangered.” He was far from alone in this fear.
Madison’s most enduring intellectual contribution to the debate over ratification is his challenging argument that individual liberties, such as property rights and freedom of religion, are better protected at the national than the state level. His argument, greatly simplified, is that the most serious threat to individual liberty is the tyranny of a majority faction. Since any given faction is more likely to be concentrated in a particular locality, and to be no more than a small minority in the nation as a whole, it follows that factional tyranny is more likely in the state legislatures than in the Congress of the United States. This argument is supplemented by others, based on the “proper structure of the Union”—deliberative representation, separation of powers, and checks and balances—that also suggest that the federal government is a superior protector of rights. Madison’s argument blunted the Anti-Federalists’ appeal to state sovereignty as the guarantor of liberty. It was, however, only partially successful. Why?
Madison’s theory gains support from robust modern social science evidence that homogeneous groups will tend to adopt policies more radical than those that individual members of the groups previously supported. Anyone who has been in a one-sided political gathering (such as a faculty meeting) will recognize the phenomenon. One-party states tend to go to unreasonable extremes. Certain states (California, Mississippi) are overwhelmingly dominated by one political party. The United States as a whole is very closely divided. Hence, the enduring plausibility of Madison’s thesis. If we are concerned about the rights of politically unpopular minorities, we should locate rights protection at the national level.
Public choice theory has, however, cast some doubt on elements of Madison’s theory. In particular, Madison’s assumption that the possibility of minority tyranny is neutralized by majority vote requirements and that minority factions are inherently vulnerable to majority tyranny is undermined by studies showing that a small, cohesive faction intensely interested in a particular outcome can exercise disproportionate influence in the political arena. If these theories are correct, Madison underestimated both the dangers of minority rule and the defensive resources of minority groups. Moreover, some observers have suggested that the conditions of modern federal politics—especially the balkanized, issue-oriented conjunction of bureaucratic agencies and committee staffs—is especially susceptible to factional politics. Political scientist Keith Whittington thus argues that decentralization may be preferred because federal politicians are too responsive to special interest groups—the modern equivalent of Madison’s “factions.”
But even taking Madison’s fundamental insight as correct—and surely it has much to commend it—the argument on its own terms cautions against total centralization of authority in Washington. It points instead to a hybrid system in which states retain a major role in the protection of individual liberties.
Madison’s argument demonstrates that factional oppression is more likely to occur in the smaller, more homogeneous jurisdictions of individual states. But it does not deny that oppression at the federal level, when it occurs, is more dangerous. The lesser likelihood must be balanced against the greater magnitude of the danger. The main reason oppression at the federal level is more dangerous is that it is more difficult to escape.
Recognition of this feature of decentralized decision making does not depend on any particular ideological understanding of the content of “liberty.” All it takes is policy diversity, which American has in spades. Some may move to avoid high taxes, some to avoid anti-transgender laws, some to escape coercion to join a union, some to be eligible for welfare, some to be able to carry guns, some to get protection from crime, some to live under more sensible pandemic regulations (whatever those may be), some to find freedom to express themselves, some to get an abortion.
Madison pointed out that there are two different and distinct dangers inherent in republican government: the “oppression of [the] . . . rulers” and the “injustice” of “one part of the society against . . . the other part.” Significantly, while Madison argued that the danger of factions is best met at the federal level (for the reasons familiar from Federalist no. 10), he conceded that the danger of self-interested representation is best tackled at the state level.
This insight strikes this author as more questionable. As an abstract proposition, it is hard to know where the danger of entrenched, unrepresentative rule is worst. The idea of a “deep state” is likely exaggerated and to a degree paranoid, but it is hard to deny that the federal bureaucracy has its own interests and commitments, which are persistent over time and largely impervious to elections. On the other hand, most big cities have been in the grip of one-party rule for decades. Local journalism, and with it the likelihood of popular accountability for city governments, has atrophied. Particular ideological and economic factions seem to dominate at both levels. Which are worse?
Public spiritedness
Critics of governmental centralization warned that public spiritedness—then called “public virtue”—could be cultivated only in a republic of small dimensions. The only substitute for public virtue was an unacceptable degree of coercion, compatible only with non-republican forms of government. There were two reasons many founders believed that a centralized government would undermine republican virtue.
First, public spiritedness is a product of participation in deliberation over the public good. If the citizens are actively engaged in the public debate, they will have more of a stake in the community. The federal government is too distant and its compass too vast to permit extensive participation by ordinary citizens in its policy formulations. By necessity, decision making will be delegated to agents. But as they are cut off from active participation in the commonwealth, the citizens will become less attached to it and more inclined to attend to their private affairs. Second, the natural sentiment of benevolence, which lies at the heart of public spiritedness, is weaker as the distance grows between the individual and the objects of benevolence.
Do these arguments still hold weight? It is a matter of contention. Are smaller towns places of public virtue and political accountability, as the Anti-Federalists thought, or of narrow-mindedness and prejudice, as Madison’s theory might suggest? We are still debating this. They are opposite sides of the coin. The very features that make smaller units of government closer to the people are also the features that make minorities within those communities uncomfortable.
We can have effective, responsive, majoritarian democracy or we can have maximal latitude for minority deviation from majority norms, but we cannot have both—except, perhaps, by the device of lodging power at one level for one kind of decision and another level for other decisions.
Whatever our chosen theory of interpretation, it is good to cast our minds back to the time of the founding, when popular attention was directed, uniquely in our history, to the issues of self-government. It is the only way to recall, and perhaps recapture, what we may have lost.
