The argument of [amazon_link id=”0691161623″ target=”_blank” ]Why Government Fails So Often and How It Can Do Better[/amazon_link] by Peter Schuck is set out wonderfully succinctly in the title, and the book does an excellent job of telling half of the story about the role of governments and markets in delivering economic outcomes.

[amazon_image id=”0691161623″ link=”true” target=”_blank” size=”medium” ]Why Government Fails So Often: And How It Can Do Better[/amazon_image]

The chapters cover a range of reasons for ‘government failure’. To list them, they are: incentives not aligned with the policy’s aims; non-rational choice; lack of information; lack of flexibility in delivering outcomes when circumstances change or things don’t work out; lack of government credibility so essential co-operation is not forthcoming; mismanagement including fraud and abuse. Schuck argues that these barriers to policy success have a “deep, structural, endemic nature.”

The book has many examples of policy failure – it’s an American and to be honest far less amusing version of [amazon_link id=”1780744056″ target=”_blank” ]The Blunders of Our Governments[/amazon_link] by Anthony King and Ivor Crewe. It’s hard to argue with the examples. This book cites also Clifford Winston’s [amazon_link id=”B00D0DIABU” target=”_blank” ]Government Failure versus Market Failure[/amazon_link], which has many more. indeed, there have been loads of policy failures, in all kinds of places and contexts.

An aspect of the argument here that I strongly agree with is the failure of policy analysts to build themselves into their ‘impact analysis’ or whatever framework they use for assessing the likely success of the initiative. In other words, the incentives the policy will create for the people affected to change their behaviour are hardly ever incorporated. Economists often think of themselves as being ‘outside’ the society, in a benign deus ex machina role.Yet all policies alter people’s behaviour and have many side-effects.

Schuck’s book does end with a chapter on policy successes – in fact it finds nine, including Airline Deregulation in 1978, the 1975 Earned Income Tax Credit, the food stamp program, the interstate highway system and the 1965 Voting Rights Act. However, it concludes: “It is hard to know for sure why these (and other) policies have succeeded when so many others have failed. Low costs, simple implementation, strong public good characteristics, and replacing far worse policies are all given as potential explanations. However, Schuck also concludes: “To succeed, the programs needed to engage the actors’ self-interest; they did not need to create new values or transform behaviors.” But he believes that the ‘low hanging fruit’ has gone.

Hence his main recommendation – be cautious. “Realistic meliorism” – make things a little bit better but keep your ambitions modest. The policy ‘doing better’ bit of the book’s title is doing far less.

I’m all for realism. There’s a missing half of the story here, though, which is how government actions unavoidably shape markets, so that to argue ‘don’t do much and just leave it to the market’ is in itself a policy. Collective choices are inevitable and government is how we make those choices. [amazon_link id=”0691161623″ target=”_blank” ]Why Government Fails So Often[/amazon_link] should be read alongside Colander and Kuper’s  recent book [amazon_link id=”0691152098″ target=”_blank” ]Complexity and the Art of Public Policy,[/amazon_link] which is about policy as determining the structure of a complex, and uncontrollable (in the old-fashioned policy sense) economy and society.

That approach is hard to get right too, but as it’s impossible not to have a structure within which markets operate, because here we are at a point in history where we have actually existing markets, it surely makes sense for governments to think about that structure. And while caution, in the face of the record of policies ranging from the inept to the horribly counter-productive, is surely sensible, thinking about structure does not automatically point to incrementalism.  Sometimes a cautious giant leap might be just the thing.