Diffusing Bad Ideas: What the Migration of the Separation of Powers Means for Comparative Law
The Separation of Powers
As an idea which was ostensibly derived from the domestic constitutional practice of a single state but which has subsequently been adopted as a core constitutional value in many jurisdictions, Montesquieu’s separation of powers model should be ‘a poster-child’ for the transplantation of constitutional innovations. Widely celebrated and globally diffused, the concept of a tripartite separation of powers is generally regarded as one of the most important constitutional insights of the Enlightenment era.
It has been cited as an influence on the development of the constitutional structures of multiple states, varying from the eighteenth century’s radical new American republic to those states that more recently obtained independence in the post-colonial era1 or aftermath of the Cold War.2 Indeed, the theory’s ubiquity is such that it even has political and normative resonance in states that do not traditionally adhere to a tripartite model of government. It was instructive, for example, that the UK’s creation of a Supreme Court and reform of the role of the Lord Chancellor were discussed in separation of powers terms,3 even though the separation of powers has not traditionally been regarded as part of the United Kingdom’s constitutional system.4 That the theory could be cited as a legitimate institutional norm in such a historically alien context underscores its broad applicability and apparent efficacy as a constitutional transplant.
However, the popularity of the separation of powers has often obscured the fact that the tripartite model suffers from several fundamental flaws. While the contention that the separation of powers is a ‘bad idea’ may be a controversial one, it is instructive to note that the theory has been subjected to sustained and consistent criticism for several decades.5 Geoffrey Marshall, for example, famously dismissed the theory many years ago as ‘infected with so much imprecision and inconsistency that it may be counted as little more than a jumbled portmanteau of arguments for policies which ought to be supported or rejected on other grounds’.6 More recently, Bruce Ackerman has commented that:
[I]t is past time to rethink Montesquieu’s holy trinity. Despite its canonical status, it is blinding us to the world-wide rise of new institutional forms that cannot be neatly categorised as legislative, judicial or executive. Although the traditional tripartite formula fails to capture their distinctive modes of operation, these new and functionally independent units are playing an increasingly important role in modern government. A “new separation of powers” is emerging in the twenty-first century. To grasp its distinctive features will require us to develop a conceptual framework containing five or six boxes – or maybe more.7
Criticisms of the separation of powers are, accordingly, not new: for that reason, it is not proposed to address this preliminary assumption at any great length in this piece. It is useful, however, briefly to identify the theory’s key weaknesses as they necessarily inform any assessment of what the diffusion of the model tells us about comparative constitutional law.
The tripartite model was ostensibly presented by Montesquieu as an overview of the workings of the English constitutional system. However, it has been persuasively argued that Montesquieu’s description of England’s institutional structures – his trivialising of the judicial power for example – was inaccurate.8
The classical version of this model prescribes a tripartite separation of functions between legislative, executive and judicial branches. In practice, this is impossible to achieve. If government is to operate effectively, institutions cannot be absolutely sealed off from each other in this fashion. Madison noted the ‘impossibility and inexpediency of avoiding any mixture of these departments’.9 The result of this inherent impracticality is that contemporary constitutional structure tends not to pursue the purest version of the theory.
The theory is hopelessly indeterminate. It requires the functions of government to be allocated amongst the three branches, but provides little or no guidance as to how this allocation should occur.
Under the formalist version of the theory, powers and functions are classified as legislative, executive or judicial and allocated accordingly. However, this approach simply begs the different question of the basis for such classifications. There are many functions and powers of government which are, in the abstract, impossible to conclusively identify as legislative, executive or judicial. As Stevens J. observed:
[T]he exercise of legislative, executive and judicial powers cannot be categorically distributed among three mutually exclusive branches of government [because] governmental power cannot always be readily characterised with only one of those three labels. On the contrary […] a particular function, like a chameleon, will often take on the aspect of the office to which it is assigned.10
The functionalist understanding of the theory accepts this difficulty but, once again, simply replaces one imprecise set of criteria with another. Functionalists seek to ensure that the allocation of powers or functions preserves the inter-institutional balance necessary to provide an appropriate level of checks and balances. But how is such balance to be measured? How is it determined?
Just as the formalist theory necessarily (and misguidedly) assumes the ability to accurately define and distinguish the three functions of the state, so the functionalist approach similarly rests on the system’s chimerical capacity to define, adopt and consistently employ the essentially indefinable notion of institutional balance.11
As a model of institutional arrangement, such indeterminacy is particularly problematic. The purpose of a constitutional model of institutional structure is to assist in the organisation and operation of government. The separation of powers is, however, incapable of providing such assistance. The theory can only be usefully invoked by those charged with making decisions on matters of institutional structure after some external values have been imported into it, be they some extrinsic (and often surreptitiously adopted) theory of government or the decision maker’s own personal views. By itself, the model functions only in the abstract. This calls into question its entire value as a constitutional principle. As Barber has observed, ‘[a] utopian constitutional theory [is] a waste of time’.12
The utility of the separation of powers theory has been further undermined by the way in which conceptions of government and of the state have changed since the eighteenth century. Government for Montesquieu was a limited rule-based activity, operated by the state’s central organs. By contrast, contemporary governance is both interventionist and decentralised in character, with power and influence exercised through an array of instruments and agencies. It is unrealistic and artificial in the extreme to suggest that the complex and multi-faceted nature of modern government can be adequately described – much less regulated – by a simple three-fold categorisation of government functions.
The criticisms outlined above indicate that the model of a tripartite separation of powers has limited, if any, value as a constitutional principle of institutional organisation. In spite of this, the model continues to be cited as a fundamental tenet of many domestic constitutional systems. The ongoing employment of the theory raises additional difficulties. One is that the theory tends to mislead rather than inform, obscuring the real basis upon which decisions about questions of institutional structure are taken. Secondly, the presence and persistence of the theory can discourage or prevent the development of alternative systems of institutional arrangement. For example, accountability in a separation of powers system is usually secured by the notional accountability of a decision maker to one of the three branches of government. In reality, however, such accountability mechanisms – such as the accountability of a minister to Parliament for the decisions of administrative bodies, or the accountability of the executive to the legislature for the execution of legislative instructions under the non-delegation doctrine – do not reflect the way in power is truly exercised in modern government. The use of such nominal safeguards provides a veneer of accountability, thereby inhibiting the development of alternative mechanisms which would provide much greater protection for the values which the separation of powers system is supposed to protect.
It is clear from this brief recital of the difficulties associated with the separation of powers that it is, in its bare tripartite form, deeply flawed. Persistent academic criticism of the model across several jurisdictions has not affected either its usage or migration across many constitutional systems. This raises questions for our understanding of comparative constitutionalism. What does the successful migration of the separation of powers theory tell us about comparative constitutional law? Why is one of the most frequently transplanted constitutional theories a model that was descriptively inaccurate when first elaborated and prescriptively inadequate whenever it is operated? Is the diffusion of this bad idea an isolated case or an illustration of a potentially broader problem? The purpose of this piece is to conduct a broad examination of these issues. Drawing on work done in other fields on the diffusion of ideas, it aims to provide some preliminary views on the possible responses to these questions.
The Diffusion of Ideas
Diffusion research combines scholarship in areas such as sociology, anthropology, psychology, marketing and communication studies to examine the process by which ideas and innovations spread. One of the core findings of this research is that good ideas do not necessarily diffuse. The reality, in fact, is that the majority of innovations fail. Studies suggest that, depending on the parameters of the particular market at issue, only between 1 per cent and 30 per cent of the many thousands of new products introduced each year succeed.13 Some of these products may have inherent flaws but others amongst them are likely to present an opportunity for genuinely beneficial innovations. For those in the latter category, however, there is no guarantee of success.
Most technologists believe that advantageous innovations will sell themselves, that the obvious benefits of a new idea will be widely realized by potential adopters, and that the innovation will diffuse rapidly. Seldom is this the case.14
Rogers uses the example of the British Navy’s efforts to control scurvy to illustrate his point. In 1601, an experiment with lemon juice by a Captain James Lancaster strongly indicated that lemons had a beneficial impact on the control of scurvy on ships. Sailors who were allocated lemon juice survived while 40 per cent of those who were not given lemon juice died of scurvy. Nevertheless, it was almost 200 years before the Navy adopted this practice and over 250 years later before the Merchant Navy followed suit.
This indicates that evidence that a practice or product is beneficial will not, of itself, ensure that it will be widely adopted. On the contrary, diffusion researchers regard the migration of ideas as a complex social process. Much of the work in this field focuses on the social mechanics of diffusion. While this research also has interesting implications for comparative constitutionalism,15 this piece concentrates on the extent to which the design and characteristics of an idea can influence how it spreads. The diffusion scholarship indicates that: ‘innovation attributes are significant factors in the diffusion of innovations’.16 In particular, it has been demonstrated that certain characteristics of innovations consistently affect the speed and extent of their diffusion:
The characteristics of an innovation, as perceived by the members of social system, determine its rate of adoption. Five attributes of innovation are: (1) relative advantage, (2) compatibility, (3) complexity, (4) trialiability, and (5) observability.17
Unlike product innovations, the latter two criteria are generally not applied to the study of idea-only innovations. In the constitutional context, for example, it is difficult to conceive of how a trial of a constitutional innovation could be conducted on a limited basis. Similarly, the results of a constitutional innovation cannot easily be subjected to consistent observation or assessment. Isolated illustrations of the principle in operation may provide some degree of observability, but such examples cannot form the basis for reliable findings. They may be wrongly attributed to other factors, or they may represent an atypical but high profile example of the relevant principle in action. Constitutional structures are too complex to allow the impact of one principle to be readily observed.
For that reason, the analysis here is confined to an examination of how the first three characteristics are said to affect the diffusion of ideas, as well as the extent, if any, to which the migration of the separation of powers corroborates the analysis of this diffusion research.
It is obvious that ideas are more likely to diffuse if they are regarded as more advantageous than the other options available. The key point here, however, is that such comparative assessments of the merits of competing ideas typically depend not on their objective strengths and weakness but rather on their perceived advantages. ‘[P]erceptions count. The individuals’ perception of the attributes of an innovation, not the attributes as classified objectively by experts or change agents, affect its rate of adoption’.18
The difficulty is that errors recur in the making of these assessments. Research has shown that the human mind is naturally predisposed to commit certain errors. One prominent example is what the literature describes as availability error where ‘decisions [are based] on the most available information and not on all the evidence’.19 This error is compounded by the unreliable way in which we frequently obtain such information. ‘Information may be available because it is widely publicized, recent, dramatic, or emotional’.20 Similarly:
[W]e often assign excessive weight to confirmatory data at the expense of contradictory data. This usually involves two mechanisms: first, we often quote anecdotal cases that support our beliefs and ignore those that don’t, and, second, we exercise “optional stopping” or “satisfying” in the pursuit of data when the early data supports our convictions, but continue the search for more data when early indicators do not support our predictions.21
One infamous recent example of how advertising can take advantage of such tendencies was the Swift Boat Veterans for Truth advertising campaign that took place during the 2004 American Presidential election. Studies of the marketing strategy adopted by those behind this campaign show that their advertising had a deliberately controversial character. This was intended to ensure that the ads received wide coverage and made a greater degree of impact on individuals. The media debate about the truth or falsity of the claims made meant the claims received far more attention than their small advertising spend would have secured alone.
Furthermore, the campaign made a significant effort in advance of the broadcast of the ads to contact and brief conservative media outlets whose antipathy to Kerry made it more likely that they would engage in the form of ‘optional stopping’ referred to above. The strategy was a conspicuous success.
While most major news outlets debunked or refuted the claims of the Swift Boat Veterans, and although only a very few of their ranks had ever actually served with Kerry in combat, their message was played and replayed throughout the national media, garnering them far more exposure than their limited budget ever could have allowed. Indeed, this was part of their overall strategy.
Regardless of the accuracy of their claims, or perhaps because of their inflammatory nature, the Swift Boat Veterans were successful in casting doubt on one of the cornerstones of Kerry’s campaign: his war record.22
Another common error of reasoning is asymmetric evaluation. This is ‘one special type of the lack of objectivity involving observing and citing occasions when an event occurred and a specific outcome resulted, and overlooking all other event/outcome combinations’.23 We are more likely to recall the high-profile performance of an innovation in an atypical situation than the many mundane occasions when it functioned in a different way. In the legal context, this may manifest itself in a focus on the way in which a principle applied in the notorious ‘hard case’ rather than in an analysis of how it functions in the majority of less dramatic situations.
The importance of perception in assessing relative advantage, and the frequently flawed nature of that perception, has a number of implications for the diffusion of new ideas. The first is that it underlines the fact that the objective merits of an innovation do not guarantee widespread adoption. As a corollary, it also means that the weakness of an innovation will not necessarily inhibit its diffusion.
In terms of the diffusion of ideas, this raises the specific problem that an idea, once adopted, may be difficult to dislodge. Individuals do not directly experience the effects of ideas on a regular basis. They thus will rarely have the incentive to assess the relative advantage of established orthodoxies. In the absence of a radical malfunctioning of the relevant system, individuals are likely to trust the ideas upon which it was originally established. In this way, trust in current suppliers poses a significant barrier to entry for new innovations.24 In turn, this confers significant first-mover advantages on those ideas that are initially accepted. Research on the diffusion of ideas in other fields suggests that the importance of this should not be underestimated.
Evidence from the pharmaceutical industry suggests that the first-mover advantage is quite substantial, be it due to reputation effects, slow information diffusion, or simply “capture” of the medical profession.25
The research also indicates that these first mover advantages are particularly pronounced where the dominant innovation or idea in question successfully obtains a large market share.
Pioneers such as Coca-Cola and Kleenex have become prototypical, occupying a unique position in the consumer’s mind. Their large market shares tend to persist because perceptions and preferences, once formed, are difficult to alter.26
The normatively cohesive character of most state’s constitutions means, of course, that the prevailing orthodoxies of that system, once initially accepted, occupy an almost monopolistic position in that order. The orthodox position of established constitutional ideas is supported by their repeated invocation and application. Competing ideas are, by definition, presumptively invalid.
Thus, once an idea-only innovation is generally accepted as advantageous, it makes it very difficult for new ideas to move from innovation to wider adoption. That any assessment of their relative advantages is also likely to be flawed further compounds this difficulty.