The Role of Economic Analysis in Public Policy
14 February 2008 at 11:17 pm Peter G. Klein 9 comments
| Peter Klein |
Here are two views on the role of economic analysis in public policy, from a passage in Robert Dodge’s biography of Thomas Schelling recounting the early days of Harvard’s Kennedy School of Government:
A tension [among the faculty] began to develop over an ideological difference between two groups. The question that brought about the division concerned the proper role of the policy analyst. Schelling’s view was the same as it had been since the Cold War, and there were other economists in the school who generally agreed. They believed that the approach to policy analysis was to begin by rationally analyzing situations, seeking to understand how things work and what outcomes would be. His idea had been to “solve the puzzle first.” Policy was something that came after understanding. Throughout his career Schelling had fought against the idea of beginning with outcomes, what he saw as looking at problems backward, and had believed that strategic analysis was required in advance to understand situations before developing public policy.
A group headed by Steve Kelman and future Secretary of Labor, Robert Reich, held a different view. This group cared about policy management as well as public policy analysis. Their main argument was that values couldn’t be separated from public policy, and the antiseptic and purely analytical approach of Schelling’s group was incomplete. Policy analysis, the Reich group felt, was to be used in determining a successful path to the goal one hoped to achieve. They believed it was necessary to acknowledge and identify openly what one was trying to achieve or affirm when carrying out a policy. . . .
There was one meeting in particular where the differences were made clear. Reich and some others were presenting papers. Reich’s paper was a challenge to cost-benefit analysis, and he used a story about building a bridge from the mainland to an island to make his point. He said one might find that cost-benefit analysis led to the conclusion that the bridge to the island should not be built, because when all the calculations were done, the cost of the bridge exceeded the expected benefits from traffic back and forth. But the analysis, Reich went on, was faulty. He said that once the bride was built, the islanders and mainlanders, instead of being separate people with nothing to do with each other, would be integrated. the forecast use of the bridge was based on surveys and demand analysis prior to its building, and would give an inaccurate measure of the bridge’s true benefit. The mainlanders and the islanders would be different people after that bridge was built, they’d be a different community. And if you believe in community, you should build the bridge, no matter what cost-benefit analysis concluded.
A professor who attended the meeting remembers what came next. “I remember [Schelling] going apoplectic. I mean, he was animated and vituperative in his rejection, and he used the word ‘fascistic.’ He said something like, ‘If you think like that, you can get any conclusion you like. You start out thinking the world would be a better place with the island and the mainland connected. You ignore the evidence on the expressed demand of the people who were there, saying they’ll be different people after the fact. Well, who are you? You just overruled those people in the interest of some vision. Well, that’s a way of thinking about acting publicly that can justify anything.”
Leaving aside questions of demand revelation and cost estimation, can someone who doesn’t understand the broken-window fallacy be qualified to be Labor Secretary?
Entry filed under: - Klein -, Classical Liberalism, Methods/Methodology/Theory of Science.
Organizations and Markets 
1.
Shawn Ritenour | 15 February 2008 at 9:58 am
Well…Robert Reich was.
2.
Shawn Ritenour | 15 February 2008 at 10:03 am
Seriously, one comment that has stuck with me from graduate school was made by Leland Yeager in his last money and banking course. When asked about the role of the economist in public policy he quickly and forthrightly declared, “The first job of the economist is the get the analysis straight.” After you get the analysis done, then you can examine what is politically possible and let the politicians do their work.
3.
Peter Klein | 15 February 2008 at 10:21 am
And, of course, technical analysis and policy advocacy should be complements, not substitutes. The claim that one should start with analysis doesn’t imply that the economist (or sociologist or political scientist or management theorist) cannot also be an advocate for particular policies or strategies, which is what some scholars seem to believe.
As Murray Rothbard put it: “[W]here is this alleged tradition of requiring economic theorists to take up the monkish cowl and abstain from all thoughts or implications of their work, let alone take direct posts in government? . . . [T]he call for political abstinence is almost always directed to economists outside the mainstream politics of the day. If economists advocate generally accepted policies, this is somehow subsumed under the rubric of “value-neutrality”; only adhering to policies opposed to the conventions of the day is decried as an intrusion of unclean political considerations into the virtuous realm of economic science.”
4.
Shawn Ritenour | 15 February 2008 at 10:56 am
Right, Ho!
5.
Rafe Champion | 15 February 2008 at 4:12 pm
Bill Hutt wrote a book (name eludes me) on the responsibilities of economists to engage with policy (on the basis of good analysis). He also wrote a book, probably in the genre of the sociology of knowledge, on the retreat from sound political economy during the nineteenth century. http://www.the-rathouse.com/Revivalist4/RC_Huttachieve.html
6.
Rafe Champion | 15 February 2008 at 4:18 pm
An extract from the piece on Hutt
One would like to see a replay of history with the economists, especially Lord Keynes, following the advice and the example of Bill Hutt. In Politically Impossible: An Essay on the Supposed Electoral Obstacles Impeding the Translation of Economic Analysis into Policy he proposed a “dual formula” for economists to deal with the problem of unpopular policies. He quoted Milton Friedman on the duty of economists to prescribe what should be done, without courting political approval. He then noted that Friedman had departed from his own advice on occasions.
“Friedman’s maxim implies that the economist’s role is to do this (the ideal) and not to do that (the expedient). I suggest it is the economist’s role and duty (in public policy discussions) to do both. Why should not advice proferred typically take the form of saying to the politicians (and indirectly to electorates) with complete candour, something like the following?”
‘In our judgement, the best you will be able to get away with is programme A along the following lines; but if you could find a convincing way of really explaining the issues to the electorate, our advice would have to be quite different. We should have to recommend programme B along the following lines’.”
“I am not suggesting that economists ought ever close their eyes to political realities. On the contrary, when they are concerned with the practical applications of their science, they ought in every instance to bring voting prospects into the picture – but explicitly.”
7.
Shawn Ritenour | 15 February 2008 at 11:38 pm
Hutt’s 1936 book is ECONOMISTS AND THE PUBLIC
8.
Fred Thompson | 16 February 2008 at 3:00 pm
It is hard to believe that Robert Reich would be right and Tom Schelling wrong. Schelling is a genius; Reich is not. Consequently, I think any confusion about the conflict between evaluation and analysis must be due to the biographer not to Schelling. Policy analysis, like management, is a design science – it’s about making things better, i.e., in Herbert Simon’s formulation, converting existing conditions into preferred conditions, thereby creating value. The idea that you can do policy analysis without talking about values is about as coherent as the sound of one hand clapping. Indeed, one of the greatest benefits of benefit-cost analysis is that gives standing to the values of folks who might otherwise not be heard in partisan, adversarial debates.
9.
Donald A. Coffin | 18 February 2008 at 1:08 pm
You know, this sounds to me like an old, old debate in welfare economics. If we’re at production set Z1, it may be Pareto optimal, in the sense that no one can gain without someone else losing. But if we’re at Z2, the same may be true. So what’s optimal? Isn’ the answer, “It depends”?
My take on the Reich/Schelling contretemps described above is that they are both wrong. Reich for assuming that the “after-the-bridge” Pareto optimum would include the bridge and Schelling for assuming that the pareto optimum coul;d never include the bridge.