For the Symposium on Michael Klarman, The Framers' Coup: The Making of the United States Constitution.

Professor Michael Klarman’s treatise on the making of the Constitution of the United States is a coup in three distinctive senses.  The Framers’ Coup is a coup in the sense of being a brilliant accomplishment.  As the blurbs and every previous review correctly note, the work is the best extant one-volume analysis of the framing.  The Framers’ Coup is a coup in the sense of being a takeover.  Klarman has clearly replaced Charles Beard and many others as the scholar whose work stands for the proposition that the Constitution is best understood as the product of an undemocratic elite who through deft maneuvering foisted a constitution on the American people that contained far more procedural and substantial protections for property holders than for the rights and interests of ordinary citizens.  The Framers’ Coup is a coup in the sense of being an difficult achievement.  Klarman has read more primary material than any other scholar writing on the framing period, mastered that material and organized that material into a remarkably accessible read.  The Framers’ Coup is the rare work that can be devoured in the library or enjoyed on the beach.

The Constitution of the United States was also a coup in the same three senses as The Framers' Coup.  The Constitution of the United States is conventionally regarded as a coup in the sense of being a brilliant accomplishment.  For much of American history, James Madison and friends were regarded as demi-gods, whose sacred text was not to be disturbed in any way.  The Constitution of the United States was a coup in the sense of being a takeover.  Charles Beard, Woody Holton and now Michael Klarman document how an alliance of bisectional elites subverted both the Articles of Confederation and the Continental Congress when fashioning a new regime in their propertied image.  The Constitution of the United States was a coup in the sense of being a difficult achievement.  William Riker and other public choice scholars point out how the framers during both the drafting convention and ratification conventions manipulated rules and political processes in order to solve the complicated and nearly intractible problems that occur when political leaders must aggregate inconsistent and multi-dimensional preferences.

Professor Klarman presents the framers’ coup primarily as a takeover, but the text of The Franers' Coup provides substantial evidence that the framer’s coup was a difficult achievement.   Americans during the framing period were badly divided by region, vocation, religion, economic status and ancestry.   They disputed paper money, access to the Mississippi River, slavery, federal courts, the structure of the national legislature, and the nature of the war power.  Klarman is right to note that the Constitution of 1787-89 was biased toward the interest of certain property holders, was not particularly majoritarian and was not ratified in a process that merited a democratic stamp of approval.  Still, the public choice literature (which my colleague Maxwell Stearns, whose helpfully reviewed a draft of this review, and many others are far more familiar than I am) suggests that every solution to the constitutional problems Americans faced during the late 1780s would be biased towards some set of interests, not particularly majoritarian and not ratified in a process that merited a democratic stamp of approval.  The Framer’s Coup suggests good reasons for thinking that constitutional preferences after the Revolutionary War did not have the structure necessary to produce a democratic outcome in the sense in which Klarman understands democracy.

The Framer’s Coup claims, somewhat in passing, that constitutional preferences at the time of the framing were structured in ways capable of generating majoritarian outcomes.  Klarman writes "opinion on the Constitution should be seen as existing along a continuous spectrum" (310),  Some Americans favored more national powers than others.  Some Americans were more democratic than others.  A second constitutional convention was an “intermediate” position between ratifying the Constitution and retaining the Articles of Confederation (545).  A median voter appears to exist whose preferences on national power, democracy and the Constitution can defeat by majority vote the preferences of any other voter.  If, for example, the median voter in 1787 preferred 20 units of national power and Americans were faced with a choice between 20 and 22 units of national power, then the half of the citizenry that preferred 20 or fewer units of national power would vote for 20 units, along with whatever fraction of citizens preferred between 20 and 21 units of national power.  On the assumption that preferences could be neatly organized from no government power to all government powers, Klarman can comfortably reach his conclusion that the Constitution of 1787-89 vested the national government with more powers than most Americans believed necessary or proper.  At the very least Klarman's claim that "the Federalists managed to secure the ratification of a constitution that was vastly different from what most Americans would have expected or wanted" (540) seems to implies the possibility of alternative constitutional arrangements that most Americans wanted..

We might nevertheless imagine a different preference ordering in which any majoritarian solution to pressing constitutional problems was impossible when the Constitution was framed.  Imagine that in 1787 one-third of all Americans wanted only an increase in national power to enforce treaties, one-third of all Americans wanted only an increase in national power to prevent states from issuing paper money, and one-third of all Americans wanted only an increase in national power to prevent trade wars between the states.  All Americans are willing to pay the taxes necessary to support other national powers if their preferred power is included in the constitution, but all Americans also prefer to have their taxes as low as constitutionally possible.  Under these conditions, no median voter exists. No constitution beats all other constitutions in a pairwise comparison.  As is well known in the public choice literature, if we engage in limitless pairwise comparisons between different possible constitutions in which citizens always vote their sincere constitutional preferences and make several other plausible assumptions, the result is a cycle rather than the choice of a constitution with particular power arrangements.

Suppose we begin with by pitting a constitution that requires people to pay taxes for a government strong enough to enforce treaties, prevent paper money, and prevent trade wars between states against a constitution that requires persons to pay taxes only for a government strong enough to enforce treaties and prevent paper money.  The second constitution will win by gaining the votes of citizens who favor treaties and oppose paper money.  We then pit that constitution against a constitution that requires persons to pay taxes only for a government strong enough to enforce treaties.  The latter constitution will win by gaining the votes of citizens who favor treaty powers and citizens who favor powers to prevent trade wars (who will see their tax burden decreased).  We then pit this winning constitution against the old constitution with no increased powers.  That constitution will win by gaining the votes those citizens who favor national power to prevent paper money and national power to prevent trade wars. We then propose the constitution with all three powers.  That one wins unanimously, since, as noted above, everyone prefers a constitution with their preferred power, even if that means paying taxes for less desirable powers.  In short, constitutions with two powers defeat constitutions with all three powers, constitutions with one power defeat constitutions with two powers, the constitution with no power defeats the constitutions with one power, but the constitution with three powers defeat the constitution with no powers.  

This preference ordering takes some sting out of Professor Klarman’s conclusion that the Constitution of 1787-89 contained more powers than most Americans wanted and was not ratified by particularly democratic procedures.  When a constitutional people have inconsistent, multi-dimensional preferences, no constitutional arrangement will be majoritarian in the sense of being preferred by a majority of people to all other possible constitutional arrangements.  Under the above preference ordering, no constitution can contain the combination of powers most people prefer, because every constitution can be defeated by some other constitution in a majority vote.  Moreover, as the late great Kenneth Arrow brilliantly demonstrated, the only way to produce a result when people have inconsistent, multiple dimensions is to violate some basic democratic norm.  We can get a stable result in the above hypothetical, for example, if we agree that any constitution defeated in a pairwise comparison may not be voted on again and also agree on the order in which we will vote. Significantly whoever gets to impose the voting order will often be able to structure alternatives so that their constitutional wins.  The framers, for example, eased the ratification of the Constitution of the United States by ensuring that the vote that mattered would be the Constitution with lots of powers against the Constitution of hardly any powers.  Change the alternatives (or the order in which alternatives were voted on, and the result is likely to change. 

Whether the substantial overlap that existed among the framers with respect to their preferences for different government powers was sufficient to overcome the public choice problems noted above is doubtful.  Substantial overlap clearly existed.  Most framers wanted a government with substantially increased powers to enforce treaties, prohibit paper money and prevent trade wars between the states.  Still, given the numerous powers the national government assumed after 1789 and the significant differences between the main participants in debates over war powers, treaty powers, commercial powers, and powers to prevent paper money, whether Americans preferences in 1787 about government powers could be placed on the single continuous dimension that easily identifies a median voter requires more evidence that Professor Klarman provides in The Framers’ Coup.

The main thesis of The Framers’ Coup still stands, even if in need of some modification.  The framers did produce a constitution with more powers than most Americans wanted.  Even if the framers could not have produced a constitution with national powers consistent with the preferences of most Americans, that the constitution they produced was biased towards one set of interests rather than another remains significant.  Given the probable structure of constitutional preferences in 1787-89, any constitution produced would have been biased towards some set of interests and generated in a way that violated certain established norms.  We can debate the interests the framers chose to protect and the means they used, but we should probably avoid the mistaken believe that a constitution existed that would not have been biased to some interests and which could have been ratified by purely democratic procedures.

The Framers' Coup illustrates Mancur Olson’s famous observation that propertied elites are typically better able to overcome public choice problems than more ordinary citizens.  James Madison, Alexander Hamilton and friends had better lines of communication, more resources and superior organization than Patrick Henry and other anti-Federalists.  Klarman might be understood in this sense as writing a primer on the means by which the most fortunate Americans are able to overcome cycling problems that bedevil their less fortunate rivals.  Those anxious to repeat the framers’ experience and engage in new constitutional conventions might remember Klarman’s demonstration that constitutional conventions almost always present public choice problems best overcome by the most fortunate citizens and, perhaps also, his concluding observation that less fortunate citizens fared much better when voting under the Constitution of 1789 than they did at either the drafting or ratification conventions.