Wednesday, November 5, 2014

Public Works, Economic Stabilization and Cost-Benefit Sophistry

I. Public Works and Economic Stabilization

This is where it all began. The National Resources Board's 1934 Report on National Planning and Public Works contained a radically different vision of the methods and purposes of conducting a cost benefit analysis than what has subsequently become the convention. This has profound conceptual (and possibly legal?) consequences for the supposed "economic optimization" of action to limit climate change.

John Maurice Clark was the NRB's economic consultant on the issue of "the use of public works as an economic stabilizing device." His findings provided the substantial basis for the report's Section II, Part 3 "Public Works and 'Economic Stabilization.'" A comprehensive report by Clark, Economics of Planning Public Works, was published in 1935 by the National Planning Board of the Federal Emergency Administration of Public\Works.

In Chapter Nine of his 1935 book, Clark introduced the (Kahn-) Keynesian multiplier into American economic discourse. This theoretical analysis provided a rationale for including the extended "secondary effects" of work relief in the calculation of project benefits. As the NRB report had noted: 
A second series of questions involves the relations of public works to economic stabilization and the emergency problem of work relief. What part can public works play in meeting the problem of business cycles and how far can these works be made an instrument for recovery?
This "second series of questions" was given a very high priority indeed by the Roosevelt administration in the context of the Great Depression. But for Clark the employment of labor that would otherwise have been idle was more than simply a secondary benefit of public works. It was the redress of a cost-shifting "externality" that resulted from the treatment of labor by employers as a variable cost that could be dispensed with during times of business slack. 

In his Studies in the Economics of Overhead Costs, Clark (1923) had argued that labor should be considered as an overhead cost of doing business rather than as a variable cost of the employing firm because the cost of maintaining the worker and his or her family "in good stead" has to be borne by someone whether or not that worker is employed:
If all industry were integrated and owned by workers, what would be the relation of constant to variable expense? would be clear to worker-owners that the real cost of labor could not be materially reduced by unemployment.
One might argue that in a democracy, public works can be regarded as "integrated and owned by workers" and thus capable of restoring payment for the real cost of labor, if not by private employers then by the government -- which could then recover the outlay through taxation. Nor should it be assumed that Clark's attitude of reparation was not shared by the National Resources Board. The opening paragraphs of the report's foreword proclaimed in populist prose:
The natural resources of America are the heritage of the whole Nation and should be conserved and utilized for the benefit of all of our people. Our national democracy is built upon the principle that the gains of our civilization are essentially mass gains and should be administered for the benefit of the many rather than the few; our priceless resources of soil, water, minerals are for the service of the American people, for the promotion of the welfare and well-being of all citizens. The present study of our natural resources is carried through in this spirit and with a desire to make this principle a living fact in America. 
Unfortunately this principle has not always been followed even when declared; on the contrary, there has been tragic waste and loss of resources and human labor, and widespread spoliation and misuse of the natural wealth of the many by the few. [emphasis added]
The conservation movement begun a quarter of a century ago marked the beginning of an organized national effort to protect and develop these assets; and this national policy was aided in many instances by the individual States. To some extent the shameful waste of timber, oil, soil, and minerals has been halted, although with terrible exceptions where ignorance, inattention, or greed has devastated our heritage almost beyond belief.
So this, then, is the founding rationale for cost-benefit analysis, as different from today's market-appeasing conventions as chalk from cheese. And -- oh, yes -- it is THE LAW:
"...if the benefits to whomsoever they may accrue are in excess of the estimated costs, and if the lives and social security of people are otherwise adversely affected." --  Title 33 U.S. Code § 701a - Declaration of policy of the Flood Control Act of 1936
Refugees from the "1000-year flood" of the Mississippi River in 1937.

II. The Fallacy of Maximizing Net Returns

In the early 1950s, the mandate for giving prominent consideration to secondary benefits was effectively expurgated from federal government cost-benefit guidelines. The purge was carried out through two documents: the "Green Book," Proposed Practices for Economic Analysis of River Basin Projects, published in May 1950 and Budget Circular A-47 issued on December 31, 1952. Maynard Hufschmidt's (2000) chronicle of "Benefit-Cost Analysis 1933 - 1985" provides a useful overview of the sequence of events. Hufschmidt worked for the National Resources Planning Board, the Bureau of the Budget, and the Department of the Interior between 1941 and 1954.

Hufschmidt recounted that the Green Book's "treatment of the thorny issue of secondary benefits was at odds with the  practice of the Bureau of Reclamation" and it recommended that benefits "should be measured from the strict national economic efficiency point of view" rather than from the perspective of local or regional benefits. This controversial recommendation was not implemented by the concerned agencies.

John Maurice Clark was called upon again, along with two other economists, to adjudicate the issues in dispute between water resources agencies and the interagency subcommittee on benefits and costs. According to Hufschmidt, the panel of economists "recommended a cautious approach to including secondary benefits," which included separate reporting of primary and secondary benefits but did not rule out their use. [I have requested the Report of Panel of Consultants on Secondary or Indirect Benefits of Water-Use Projects through interlibrary loan and hope to elaborate on its analysis when I have had a chance to study it.]

The economic consultants' report was submitted at the end of June, 1952. Six months later it became a moot point as the federal Bureau of the Budget issued Budget Circular A-47, severely restricting the use of secondary benefits. Hufschmidt described A-47 as a "conservative document" that was regarded as imposing "severe restraint" on water projects:
The subject matter coverage was much the same as the Green Book; basically, it was a conservative document, which placed primary emphasis on economic efficiency-oriented primary benefits for project justification. The use of secondary benefits was severely restricted, an opportunity-cost concept of interest or discount rate, tied to the interest rate of long-term government bonds, was adopted, and a 50-year time horizon was established.  
Budget Circular A-47 was widely regarded by the water resources agencies and by the many proponents of water resources projects in Congress as a severe restraint on water projects. It served this purpose during the eight years of a relatively conservative Republican administration under President Eisenhower from 1952 to 1960, and was finally rescinded in 1962 in the early days of President Kennedy’s administration.
It doesn't need to be assumed that skepticism or caution regarding the evaluation of secondary benefits was unwarranted. Richard Hammond (1966) observed that the practices of the Bureau of Reclamation "brought benefit-cost analysis into disrepute in many quarters, particularly when agencies continued, in times of wartime boom and post-war 'full employment,' practices generated by the depression." On the other hand, the remedy pursued by the Green Book had its own problems, characterized by Hammond as "The Fallacy of Maximizing Net Returns" which the Green Book pursued as an "incontrovertible proposition":
The most effective use of economic resources is made if they are utilized in such a way that the amount by which benefits exceed costs is at a maximum rather than in such a way as to produce a maximum benefit-cost ratio or on some other basis... This criterion of maximising net benefits is a fundamental requirement for economic justification of a project. [emphasis added by Hammond]
"What seems to have happened," Hammond observed of the foregoing paragraph,"is that a familiar abstract proposition of economic theory, that rational conduct consists in balancing marginal cost against marginal gain, has been mistaken for a prescriptive rule of behavior applicable in any and all circumstances without qualification." Furthermore, he eventually explained, the maximizing mania ultimately boils down to substituting guesswork about one set of "opportunity cost" intangibles for other intangibles called "secondary benefit" and diminishing some of the speculative figures to an infinitesimal amount by the application of an arbitrary discount rate.

In defence of Clark's earlier formulations regarding secondary benefits, he was almost exasperating in his insistent qualification of cost and benefit estimates as judgemental and tentative. This contrasts with the maximalist language of pseudo-scientific precision exemplified by the Green Book's use of "words like measure, ascertain, and evaluate in contexts where estimate, expect, and guess would be more appropriate."

III. Kapp and Trade

In 2010 the U.S government's Interagency Working Group on Social Cost of Carbon (IAWG) presented its estimate of the social cost of carbon "to allow agencies to incorporate the social benefits of reducing carbon dioxide (CO2) emissions into cost-benefit analyses of regulatory actions that have small, or 'marginal,' impacts on cumulative global emissions." The IAWG's central estimate for the social cost of CO2 in 2010 was $21 in 2007 dollars, based on a 3% discount rate. One of the damages associated with an increased increment of carbon emissions in a given year is specified as "property damages from increased flood risk." Would the Flood Control Act of 1936 have any pertinence to their cost benefit analysis?

The Kaldor-Hicks compensation test constitutes a guiding principle for the selection of a discount rate for cost-benefit analysis, the IAWG report explains:
One theoretical foundation for the cost-benefit analyses in which the social cost of carbon will be used— the Kaldor-Hicks potential-compensation test—also suggests that market rates should be used to discount future benefits and costs, because it is the market interest rate that would govern the returns potentially set aside today to compensate future individuals for climate damages that they bear.
The Kaldor-Hicks test presumably allows the analyst to set equity considerations aside while evaluating the economic efficiency. Does it?

David Ellerman argues that the efficiency/equity distinction is simply an artifact of the choice of numeraire. In other words, the supposed efficiency of a policy outcome measured in dollars is an illusion created by the fact that efficiency is being measured with the "same yardstick" that was used to assign "value" to incommensurable things like human life, output of goods and services and damage to the environment. If one reverses the process and establishes human life or environmental damage as the unit of measurement, then the results of the analysis are also reversed.

Although simple, this is not an intuitively obvious argument, so Ellerman illustrates it with a very simple example in which John values apples at one dollar each, while Mary values them at 50 cents. Social wealth would be improved if Mary sells an apple to John for 75 cents. Under the Kaldor-Hicks criterion, social wealth would also be improved if Mary lost her apple and John found it, even though Mary receives no compensation. Kaldor-Hicks would deem this an efficiency gain because John could potentially compensate Mary by paying her 75 cents for the lost apple. Measured in apples, though, there has been no change in total wealth because Mary's lost apple exactly balances John found one..

But using apples as the unit of measurement changes everything. Since John values one apple at one dollar, he also values one dollar at one apple. Mary values a dollar at two apples.Measured in apples, social wealth would be improved if John lost a dollar -- worth only one apple to him -- and Mary, who values the dollar at two apples, found it. John's cost is smaller -- in apples -- than Mary's benefit. But since a dollar is a dollar, if the unit of measurement was dollars, the cost and the benefit would exactly balance leaving no net gain.

Ellerman's illustration may seem trivial but the "same yardstick" argument comes from Paul Samuelson who pointed out that, measured in money, the marginal utility of income is constant at unity. Bill Gates would value an extra $20 a week of income as much as a Walmart clerk would -- $20 dollars worth! It's a tautology.

Of course that's not the only problem with the IAWG's cost of carbon estimate. Moyer, Woolley, Glotter and Weisbach argued that the social cost of carbon estimates in the IAWG models are constrained by shared assumptions of persistent economic growth. Even a modest negative impact on productivity, they find, would increase social cost of carbon estimates by several orders of magnitude above the IAWG estimates.

Johnson and Hope found that assigning equity weights to damages in regions with lower incomes or using different discount rates generates social cost of carbon estimates two and a half to twelve times those of IAWG. Foley, Rezai and Taylor argued that the social cost of carbon and the relevant social discount rate are conditional on a specific policy scenario "the details of which must be made explicit for the estimate to be meaningful." There is also Martin Weitzman's analysis that the uncertainty about the prospect of catastrophic climate outcomes renders traditional cost-benefit analysis irrelevant.

Remember how we got to this analytical impasse, though? Clark's analysis of planning for public works and the National Resources Board report were concerned with the environment to be sure. But their sense of urgency was more particularly focused on the unemployment crisis. Controlling floods, reclaiming eroded agricultural land and replanting forests were viewed as ways to productively employ workers who would otherwise have to be given welfare or work at "leaf raking" make-work jobs. Public works were being considered as a way to smooth out the fluctuations of the business cycle and ameliorate the effects of cost-shifting due to employers accounting for workers as a variable, rather than a fixed overhead cost.

In February 2010, the U.S. official unemployment rate was 9.8%. The word "unemployment" doesn't appear in the 50-page IAWG report on the social cost of carbon. Nor do the words "recession," "jobs," "poverty" or "inequality" The word "labor" occurs several times but only in the context of an arcane footnote about "a method of estimating η using data on labor supply behavior." The "lives and social security of people" is given short shrift. "Growth," however, appears 34 times, about two-thirds of which refer to economic growth. In the IAWG report, one may conclude, economic growth is unrelated to employment of labor but closely correlated with "interest rate," which appears 20 times -- roughly the same frequency as "growth" in the economic context..

That's the problem right there.

In 1950, the same year the Green Book was curbing the use of secondary benefits in cost benefit analysis, Karl William Kapp's book The Social Cost of Private Enterprise was published, inspired by and elaborating on J. M. Clark's analysis of cost shifting. "As Kapp implied," remarked Joan Martinez-Alier, "from a business point of view, externalities are not so much market failures as cost-shifting successes." From that perspective, the IAWG's $21 a ton estimate of the social cost of carbon dioxide also may be better understood as an agenda-shifting success rather than a planning failure.

Eighty years ago, it may still have been possible to believe that those cost-shifting successes of business could be remedied through planning and public works conducted by a democratically-responsive government. Today, the role of government and the intention of cost benefit analysis is very different from what was professed in the foreword to the National Resources Board's 1934 report. How has that happened?
"Many difficult conceptual issues such as externalities, consumer surplus, opportunity costs, and secondary benefits that had troubled earlier practitioners were resolved and other unresolved issues, such as the discount rate, were at least clarified." -- Maynard M. Hufschmidt, "Benefit-Cost Analysis 1933-1985"
Just what are those "secondary benefits"? What are the "opportunity costs"? How did the difficult issues get resolved? And who cares?

What if I told you that "secondary benefits" was a cipher for "wages of labor," that "opportunity costs" was code for "return on capital investment" and that a significant portion of those supposedly sacrosanct financial "opportunities" result from cost-shifting? What if I pointed out that the "difficult issues" were "resolved" by declaring that wages were of little concern to public policy making but that profits were paramount?

Would you conclude with Hufschmidt that these difficult "conceptual" issues had been "resolved" or "at least clarified"? Or would you object that these are political, not conceptual, issues and that they have been suppressed and arrogated by the ideological framing and technocratic jargon of cost-benefit analysis? I leave the last word to John Maurice Clark:
"It comes down to this, that any use of labor that is worth anything at all is worth that much more than nothing. In that respect the socialist view of business depressions is correct and any rebuttal that attempts to explain away this fact by the reckonings of financial expenses is a bit of economic sophistry." Studies in the Economics of Overhead Costs (1923).


Unknown said...

the use of an interest rate was a way to arbitrate between competing uses of the funds over time

which only occurs when a government is facing full employment, which is why it emerged when it did

this was the point of Circular A-47, which was to unify the assorted plans of the water-related agencies; therefore it had to weigh discount rates.

it should be observed that in any case the federal government largely ignored A-47 until the civil rights movement reignited the legitimacy of contesting the principle of federal water resource infrastructure, and the process of public consultation caused the length of planning-permission procedures to explode.

Thornton Hall said...

Do the practitioners know about the embedded choices? One gets the feeling that it is possible to obtain a PhD and not know. It's like never traveling to a foreign country: you'd never know that much of what we consider "normal" is actually uniquely (bizarrely) American.

W. Never visited a foreign country (other than Tiajuana).

Peter Dorman said...

A couple of quick points, S-man.

First, the distinction between micro and macro analysis of public projects is tricky. Yes, if a project generates employment and increases incomes in the context of an output gap, that needs to be considered. OTOH, to apply that analysis to a single project implies that the public budget is perfectly flexible: if the project is not undertaken that much less money will be spent. But if you had a fixed budget, the size of the budget would be the relevant macro variable and you could ignore macro impacts at the project level. In the real world we're usually somewhere in the middle, right? Incidentally, there has been a big move in developing countries toward the adoption of employment-generating public programs, the biggest being India's National Rural Employment Guarantee.

Second, the Ramsey equation for a supposedly ideal discount rate assumes, as you point out, perpetual growth. It is not unreasonable to suppose that climate change could cause a substantial reduction in economic welfare, however, as it kicks into high gear. In that case, Ramsey would have us use a negative discount rate: future costs would count for more today because future people would have lower incomes and a higher marginal utility of money.

Personally, I'm not persuaded by Ramsey overall, so I don't get stressed out by this reversal. It's an interesting argument though.

Sandwichman said...

"In the real world we're usually somewhere in the middle, right?"


In what sense could a global climate-economy assessment model be rationalized as "microeconomic"?

What I'm trying to convey here is that the conventional cost-benefit analysis at the core of DICEy IAMs is a hybrid whose parts don't add up to a coherent whole, and certainly not one relevant to the issues it is called upon to address.

Thornton Hall said...

The extent to which public budgets are flexible is an empirical question that economics should have a very well known answer to. It's as if in chemistry somebody said, "Water boils around 200 degrees, right?"

"Personally, I'm not pursueded by Ramsey overall"
is a very roundabout way of claiming to know how humans will react to climate change for the next 50 years.

Unknown said...

You might find this interesting:

Sandwichman said...

Thanks, Greg, Looks very interesting indeed!

"Project Appraisal for the Keynesian Investment Planner"

Just Wondering said...

"Although simple, this is not an intuitively obvious argument, so Ellerman illustrates it with a very simple example in which John values apples at one dollar each, while Mary values them at 50 cents. Social wealth would be improved if Mary sells an apple to John for 75 cents. Under the Kaldor-Hicks criterion, social wealth would also be improved if Mary lost her apple and John found it, even though Mary receives no compensation. Kaldor-Hicks would deem this an efficiency gain because John could potentially compensate Mary by paying her 75 cents for the lost apple. Measured in apples, though, there has been no change in total wealth because Mary's lost apple exactly balances John found one..

"But using apples as the unit of measurement changes everything. Since John values one apple at one dollar, he also values one dollar at one apple. Mary values a dollar at two apples.Measured in apples, social wealth would be improved if John lost a dollar -- worth only one apple to him -- and Mary, who values the dollar at two apples, found it. John's cost is smaller -- in apples -- than Mary's benefit. But since a dollar is a dollar, if the unit of measurement was dollars, the cost and the benefit would exactly balance leaving no net gain."

I don't get this. I would think that the price of apples would play a role here. How does this work our if the market price of apples is 25 cents? What am I missing?

Sandwichman said...

Ellerman is not talking about market prices but about preferences for one good in terms of another good (including money as a good). John and Mary have difference preferences for apples relative to money.

David Ellerman said...

The apple-dollars example is from David Friedman's book "Law's Order" (p. 20). Consider a transfer of an apple from Mary to John and a transfer of $0.75 from John to Mary. Use Kaldor-Hicks to evaluate each part as a "project" with the other part as the "compensation". Using money as the numeraire and the apple transfer as the "project", we see under the assumptions that the transfer of the apple increases social wealth measured in dollars so that is the recommendation based on "efficiency", and the payment of the "compensation" of $0.75 is a matter of "equity" of concern to politician, theologians, and philosophers but not to the professional economist. Now reverse the numeraire taking apples as the numeraire and the transfer of the $0.75 as the "project". Then the transfer of the apple (= "compensation") does not change social wealth = size of the apple pie, but the transfer of the $0.75 increases the size of the social apple pie by 3/4 of an apple so it is the transfer of the $0.75 that is recommended on efficiency grounds by hard-nosed economists while the transfer of the apple is left to politicians, theologians, and the like as a matter of "equity." Thus the outcome of the KH analysis is reversed by a change in the numeraire used to describe the exact same pair of transfers. See the paper for details which can be downloaded here: