The paper by Davis, Murphy, and Topel is actually a pastiche of ridiculously obvious manipulations. I cannot in good conscience discuss it seriously at all, since it abandons any pretensions to realism or verisimilitude. For example, the authors go through the pretense of estimating the net present value of 33 years worth of containment at at average of 19% annually. Readers ought to understand that they want as low a discount rate as possible to get as high a net present value. One could reasonably expect this to be a straightforward matter, but the authors select the absurdly low figure of 2%, explaining that "Two percent per year is a reasonable figure for the real rate of interest facing the U.S. government." This makes $19 billion spent over 33 years $472 billion. Of course, if you use the more plausible rate of 4%, as Stiglitz does (remember, Stiglitz uses his rate to estimate the net-present value for debt; so he too would like a low discount rate), then the cost falls to $356; if you also include the logically consistant duration of 22 years instead of 33, then the cost is $289.² In fact, using standard OMB procedure, Iraq could have been contained in perpetuity for a mere $272 billion.

But the biggest flaw of all in the essay is a logical fallacy known as "affirming the antecedent": prove your point by inserting an assumption that the thing you seek to prove is true. The object of the paper is to prove it is proper to invade Iraq; it maintains as a premise that prior Western actions towards Iraq were needful. Since that was the very thing under debate, there is no reason to read further. I did, and it just gets worse.

Please note the predictible stance taken by the University of Chicago: this institution, famous for espousing hard-line anarcho-capitalist ideologues, could plausibly be expected to be critical of either the huge costs associated with a big government program (regime change), or else, the maximalist-state ambitions of securing national security in the most costly way possible. Initially, the Cato Institute published papers critical of the war on the grounds that it would usher in a new era of big government, while libertarians like former Rep. Ron Paul (R-TX) attacked the idea of the US government taking on the utopian ambition of spreading democracy.

The University of Chicago, however, adopted the opposite position: now that a rightwing government wished to do something huge, costly, and economically interventionist, it scrambled to support it. The usual libertarian position, that using the government as the risk manager of last resort was an inherently illegitimate and dangerous act, was sidelined. The usual libertarian position that prudential benefits of government action are not worth the precedent or the market distortions, was abandoned.

Estimates by Nordhaus (Chapter 3, War with Iraq: Costs, Consequences, and Alternatives; text, PDF) were made on the eve of the invasion and have been the most prescient. Early on he mentions an estimate made by Abbas Alnasrawi (p.7) that the 1st Gulf War destroyed $230 billion worth of infrastructure. This is an astonishing statistic: in 1991, the population of Iraq was about 18 million, so this translates to about $12,700 worth of infrastructure per capita destroyed. Prior to the Iran-Iraq War, Iraq's per capita GDP was about $2143 per capita [*], which would translate to about $84 billion or so in total fixed capital for a highly industrialized nation (which Iraq was not). Since, spending on infrastructure has been notoriously poor, while the population growth has been among the world's highest. UN/World Bank estimates of urgently-needed infrastructure spending was about $56 billion, or about $2K per Iraqi citizen.³ Generally speaking, this would be consistant with sustaining or creating a productive capacity of $500 per person per year. Even assuming the media is willfully refusing to report "all of the good news" in Iraq, 24 months of occupation has involved an expenditure of $24 billion in reconstruction, of which 13.7 has actually been spent on reconstruction (GAO, Feb '06, p.2).

It's bizarre to read GAO papers, partly because these reports contain so little useful information (none of the data mentions vanishing funds or cost accounting), and partly because the amounts involved are so small. Writing with hindsight, I know with certainty that direct expenditures by the US government on this war have exceeded $500 billion; yet, rather than ensure the Iraqi project redeems itself, the Bush Administration has cancelled numerous projects to provide water to Iraqis to save comparatively petty amounts of money:

U.S. reconstruction efforts in the water and sanitation sector focus on improving Iraq’s potable water, sewage, and sanitation systems. State [Department] reallocations have reduced available U.S. funding for improving Iraq’sseverely degraded water and sanitation sector from a peak of $4.6 billion to a current level of $2.4 billion. The United States has made some progress in completing large and small water and sanitation projects, but it is difficult to determine the impact of its reconstruction effort on this sector due to limited performance data and measures. The U.S. reconstruction program has also suffered from delays in completing projects, and some completed projects lack sufficient Iraqi staff and supplies to function properly or are not operating at all due to a lack of electricity and diesel fuel.

[...]

U.S. officials estimate that approximately 60 percent of water produced in Iraq is unaccounted for—lost to illegal taps, unmetered usage, and leaking water pipes. Because of water losses and the lack of metering, the extent to which clean potable water from improved facilities is reaching users is unknown...In commenting on our draft report, USAID said that the agency tracks increases in the amount of water treated and estimates increases in beneficiary numbers. However, these metrics do not address the quality of water and sanitation services in Iraq, ...For example, because of problems with the distribution network, water that is potable at the treatment plants may be contaminated by the time it reaches users. According to a senior PCO official in the water sector, potable water and sewage mains in Iraq are sometimes adjacent to each other, allowing leaking sewage to enter the water mains.
["Status of Funding and Reconstruction Efforts" (PDF), p.30-33]

U.S. Pays for 150 Iraqi Clinics, and Manages to Build 20 (April 30, '06-NYT)

A $243 million program led by the United States Army Corps of Engineers to build 150 health care clinics in Iraq has in some cases produced little more than empty shells of crumbling concrete and shattered bricks cemented together into uneven walls, two reports by a federal oversight office have found.

The reports, released yesterday, detail a close inspection of five of the clinics in the northern city of Kirkuk as well as a sweeping audit of the entire program, which began in March 2004 as a heavily promoted effort to improve health care for ordinary Iraqis. The reports say that none of the five clinics in Kirkuk and only 20 of the original 150 across the country will be completed without new financing.

Written by the Special Inspector General for Iraq Reconstruction, an independent office, the reports cite a wide range of factors, including disputes among Iraqi construction companies and problems with local materials, that have contributed to the program's failures. The American company Parsons, the prime contractor for the work, also comes in for stiff criticism.

[...]

Late Friday, the inspector general also released an audit report on a $147 million United States-led program to train and equip thousands of Iraqis to protect oil pipelines, electrical transmission lines and hundreds of key installations in both sectors.

Begun in September 2003, the effort, called Task Force Shield, was so disorganized that the auditors were never able to determine basic facts like how many Iraqis were trained, how many weapons were purchased and where much of the equipment ended up, the report says. Of 21,000 guards who were to be trained in protecting oil equipment, for example, probably only about 11,000 were, the report says. And of 9,792 automatic rifles purchased for those guards, auditors were able to track just 3,015.

What is puzzling is, of course, that conservatives in power sought to commit the country to a boondoggle so huge one could have attached the entire New Deal as an enabling rider. Conservative economists would logically be the most disgusted by spending per se, and dismissive of any supposed "multiplier effect" from the staggering expenditures. Conservatives, if true to character, would shake their fingers in wrath at the gigantic market distortions of such huge government outlays. The refusal to pay the piper anytime soon only increases the distortions of capital markets. The absence of accountability has been the real multiplier effect, causing this military campaign to cost almost $1 million per Iraqi. In 2003, Iraq was estimated (link above) to be economically on a par with Mauritania.

Today, of course, Pres. Bush is replaying the march to war with Iraq, now with Iran. Iran is an entirely different matter. Whereas Iraq was friendless, cut off from all forms of capital investment for 13 years, and ravaged by massive physical destruction from the first Gulf War, Iran has spent the last 10 years growing at about 7%; its GDP is now more than 30 times that of Iraq in 2003. It enjoys observer status in the Shanghai Cooperation Organization (SCO) with China, Iran, and India (IHT). Unlike Iraq, whose terrain was both flat and inhospitable to people "off the grid," Iran's terrain is extremely rugged and can sustain a large population without a fully operational infrastructure. Disturbingly, it seems that the Bush White House imagines it can bring the Iranian state to its knees without actually invading it; it seems to have convinced itself that jabs will suffice.

This administration has been most notable for the extreme determiniation it has exhibited to fortify itself against consensual reality. In interviews, time and time again, administration officials demonstrate a basic lack of either facts, proportion, or reasoning skills. The bearer of bad tidings is assumed to want the administration to fail, and somehow cause it. Defenders of the administration have been more robotic, less prone to... how shall I say this?... complete sentences. Stitching together objects and predicates, employing syllogistic reasoning, has been replaced by strings of emotive words stripped of any logical structure. And today the administration appears to think it is going to take down Ahmadinejad with a game of chicken, simply because no one has contemplated what an armed confrontation with Iran would look like.

This time around, there won't even be any rosy scenarios.

[Imperialism] [Iraq] [Economics]

Oddly, in the text, they actually say that a 3% annual PDF would mean a mean probable duration of containment of 33 years, not 22. After 33 years, there would be a 63% likelihood of a benign regime change.

2 The al-Tikriti Dynasty could conceivably rule for a long time; however, the authors arbitrarily inflated the figure after giving normal-sounding parameters. Adding 11 years and violating the principles of economics adds $183 billion to their price tag. Why not just say they expected a 2% annual PDF? Why not say they expected a Saddamist dynasty to last another 130 years (which would allow them to substitute the absurd 2% discount rate with the more conventional 4%, and their regime-change PDF to 0.53% annually)? For those of you interested in how Stiglitz came up with a discount rate of 4% (which I regard as still too low), here's the answer: in International Finance and Open-Economy Macroeconomics (Giancarlo Gandolfo, 2002, p.298), we are advised that "the marginal rate of time preference must be equal to the interest rate, since the latter reflects the relative market price of [future consumption] relative to [present consumption]." For the 10-yr inflation-indexed US Treasury, this was about 2% (10-year rate). But human beings are paying the taxes; it's the average taxpayer, paying off a mortgage, whose discount rate matters. For that person, the real rate of interest on long-term debt is >4%. The Office of Management and Budget (OMB) uses a rate of 7%, and tests for rates of 5%-9%. Using this rate, the Ba'thist regime could have been contained for all eternity for about $272 billion.

So, alright: why is using the mean 10-year inflation-indexed US bonds not valid? The answer is that it is arbitrary; it's the bottom end of a hierarchy of interest rates paid by institutions. The real interest rate paid by the US government for existing debt has nothing to do with the discount rate, and varies depending on Fed policy. Part of the reason it's so low is that government paper is something banks have to carry, and something they're handsomely rewarded for carrying. Setting the discounted utility rate at this level ignores the fact that humans actually arbitrage money and time at rates midway up this hierarchy, with credit card debt or loansharking at the top.