John McGinnis certainly doesn’t underestimate the importance of his task in Accelerating Democracy: Transforming Governance Through Technology. By page 4, he makes clear that “mass disorientation” caused by rapid technological change “can become the source of both national aggression and non-state terrorism… the dynamic of modern technology could as easily lead to a nightfall of civilization as to the dawn of a far better world.”

With that bold statement of the stakes, McGinnis, the George C. Dix Professor of Constitutional Law at Northwestern University, takes on the very Madisonian task of considering the design of a government for the twenty-first century (to say “and beyond” would almost seem to ignore an important component McGinnis’s thesis: that technological change is coming so rapidly that tremendous and ongoing flexibility in government structures will be required).

For purposes of this review, we can breeze over some of the most interesting reading in Accelerating Democracy. McGinnis devotes the opening chapters to reviewing the rapid and accelerating nature of technological change, and argues convincingly that such change creates upheavals that can lead to mass social unrest and disruption. Additionally, technology, while opening new avenues for progress, can also pose greater threats. To take an easy example, the advent of the nuclear age offers a great source of clean energy while creating a waste disposal problem rarely if ever seen before – not to mention the destructive power of weapons resulting from some of the same discoveries.

Meeting this challenge requires updating government structures to assure better public decision-making, and here we find the meat of McGinnis’s project. As McGinnis sees it, the challenge is to improve government decision-making by finding new ways to take advantage of dispersed knowledge recognized by Hayek, and using the accumulated data and knowledge that results to develop better solutions and overcome biases that prevent sound public policy.

With this in mind, McGinnis urges government not to stand in the way of developments in artificial intelligence, seeing such efforts as futile, and arguing that the dangers of AI are more than offset by the potential gains. For the most part, however, McGinnis’s recommendations are remarkably low-tech. Other than a tremendous confidence in prediction markets (such as Intrade) to aggregate knowledge into a distillable form that can be used to predict the results of various policy options, and thus create better public policy, many of McGinnis’s recommendations are simply low-tech policy changes aimed at assuring that current public policy does not cut off the flow of information to voters and government decision-makers. Accordingly, he is critical of campaign finance laws that reduce the flow of information to voters; he urges government policies that encourage new and dispersed media (such as providing bloggers with some of the same protections now available by statute to old media); and he would have government subsidize prediction markets, internet access, and, it appears, political campaigns. He wants to end political gerrymandering, which he sees interfering with sound policy by emphasizing ideology over pragmatism in both voters and their representatives. He favors term limits on representatives, mainly to assure younger, more tech savvy lawmakers, and to erode status-quo bias.

All of this is good insofar as it goes. I wish, however, that McGinnis, a first rate mind, had gone much further in rethinking the status quo himself. McGinnis, who is hardly a big government acolyte, accepts at the outset that government is big and going to stay that way, and fashions his recommendations within that framework. Indeed, he argues that the modern administrative state is “a response to technological change.” There is some truth to that, of course, but we cannot forget the modern state is also a response to a change in political philosophy, away from classical liberal ideals and towards progressivism. Huge parts of the administrative state have little if anything to do with technological advancements. The entitlement state and massive programs of redistribution and social insurance reflect philosophy more than changing technologies. Wage and hour regulation and anti-discrimination laws aimed at promoting fairness, and the tremendous bureaucracy they require in both government and the private sector, are less about technology than ideology and understandings of the role of government. Wage and price controls, funding for the arts, and bailouts of favored industries create administrative bureaucracies unrelated to technological advancement. The list could go on at length.

By accepting the existing roles of the state (and at several points arguing that greater government subsidies in new areas will help it run better), McGinnis dramatically limits his options. Classical liberal politics and economics, with their emphasis on individual freedom and decentralization, were extremely effective at aggregating dispersed knowledge through pricing mechanisms, allowing for flexible public policy, and minimizing the harms of public policy mistakes created by imperfect information and bias. McGinnis implicitly acknowledges this, and makes nods to federalism and decentralization at times.  But Accelerating Democracy does not undertake a fundamental rethinking of the administrative state, or assess how new technologies might allow a more minimal state to cope with the types of problems that were perceived to demand a larger state in the past. If its more efficient, better government we want, certainly the possibility that government must be smaller, not as a matter of particular political decisions but as a matter of principle, might be a good place to start.

McGinnis dismisses such ideological struggles as “value arguments.” He seeks, instead, to design better public policy for “valence issues,” that is, issues that everyone, or at least most people, agree on. For example, few and far between are those who do not want, at some level, education, or a clean environment, or a better standard of living. McGinnis focuses on “the best means to reach [such] common goals.”

But even valence issues do not usually exist in a vacuum. Education is indeed a valence issue, but so, for example, are low taxes.  Public policy disputes exist exactly where these two valence issues come into conflict. Of course, it may be that better information on policy outcomes will harmonize, to at least some extent, those issues, by allowing voters to improve education at minimal cost.  But at some point, a trade-off still has to be made. Similarly, tradeoffs must be made between valence issues such as environmental protection and economic growth, or government regulation in any number of areas, and personal freedom. McGinnis believes that given a better process for assessing the results of past policies and the probable outcome of future policies, it may well be that classical liberal ideals will be shown to be superior and thus rise to the fore with both the electorate and policy-makers.  Perhaps, but the Founders sought to install a framework to assure freedom, not merely to give the populace a means to evaluate the effectiveness of policy.

McGinnis also may over-emphasize democracy as a means of “assess[ing] and predict[ing] the consequences of social policies.“ Whatever else its benefits – and there are, of course, many – this is probably not democracy’s strongest point. McGinnis is aware of and cites to some of the criticism of electoral ignorance that leads to bad policy, but dismisses that largely on the grounds – quite correct, by the way – that we still ought to develop mechanisms that help democracy get it right. But it may well be freedom, more than democracy, that creates good results. Despotic democracies, such as that in Chavez’s Venezuela, can be quite as bad as many benign dictatorships.

For the classical liberal, it is freedom, not democracy, that generally leads to sound policies, and this seems reasonably well demonstrated in history.  The purpose of democracy is to protect freedom. Yet assaults on freedom seem to be able to withstand numerous policy failures, even in democratic countries, and it is not at all clear that this is merely because of an inability to evaluate policy outcomes for success or failure. As Jefferson famously noted, “the natural progress of things is for liberty to yield and government to gain ground,” – the benefits of liberty notwithstanding.

This is not to say that McGinnis’s limited solutions are wrong. Most I agree with, some I do not. Most of them would be worth pursuing even if we were not in a period of accelerating technological advances. But even within the narrower framework of what he calls “consequentialist democracy,” McGinnis’s recommendations might have roamed further. For example, in arguing against gerrymandering, McGinnis calls for independent districting commissions. Leaving aside the merits of that idea, however, McGinnis does not mention the Voting Rights Act, which, as interpreted by the courts (and apparently with support of congressional majorities) requires the creation of “minority/majority” districts, in which minorities comprise a substantial majority of voters in the district. This tends to bleach minority voters out of some districts, while packing them into others. The result is a legislature of John Conyers and Maxine Waters types mixed with Republicans from districts with few or no minority voters at all. A more adventuresome prescription would call for rethinking of this Act.

Similarly, McGinnis would like to see reform of the primary process, which he believes will reduce partisan bias, but doesn’t challenge the idea of state-mandated primaries themselves. There are reasons to consider, however, whether primaries themselves, regardless of form, lead to the type of bias McGinnis would like to push out of the system.

In the end, then, Accelerating Democracy is a valuable book on governance. It’s lengthy discussion of prediction markets should be an important contribution to thinking about how to get public policy right, and beyond that it offers numerous ideas for improving the process of formulating government policy, only a few of which I have mentioned here. My only wish is that McGinnis had allowed himself to roam further – especially given the stakes he enumerates at the outset.