THERE ARE A large number of technical issues that we have had to address in analyzing the costs of the Iraq war. In this technical appendix, we examine several critical issues and explain some of the underlying reasons for our approach and our conclusions. It should be added that many of these issues are very complex—whole tomes have been written on them—and in this short appendix, we cannot do full justice to them.
Oil has been at the center of the war from the onset. Many believe we went to war to get an assured supply of inexpensive oil for the United States and its oil companies. We begin by explaining both why that belief is so widespread and why the quest for an assured supply of oil could, or at least should, never have been the basis of a rational strategy. But whatever our motives, the consequences have been the opposite: oil prices have soared. In the third section, we explain why we believe that the war should be given “credit” for much of the rise—and why our assumption, attributing but $5 or $10 per barrel of the rise for seven or eight years—is excessively conservative.
In the 1970s, soaring oil prices played a central role in the macroeconomic disasters of that decade. This time, so far, the effects have been more muted; we explain why this is so but also why the effects are still significant, far greater than just the effect of the hundreds of billions of dollars that have been transferred to the oil-exporting countries.
In the text, we argue that one of the effects of the soaring oil prices was to dampen the economy; had prices been lower, output would have been higher. But was there scope for expanding production? In the fourth section, we explain why there was scope, and why, more generally, had we spent the money in ways that would have stimulated the economy more (than the dollars squandered on Iraq), the economy could (and would) have been stronger.
Many of the macroeconomic effects of the war are hard to quantify—for instance, markets dislike uncertainty and the turmoil in the Middle East has clearly contributed to uncertainty. And while most of our analysis has focused on the effects of the war on aggregate demand (the amount Americans have to spend on goods at home is reduced because we are spending more on oil), there are also supply-side effects. As labor gets diverted to the war effort, as the numbers of casualties mount, and as the war has diverted resources away from needed investments in both the public and the private sector, the economy’s productive potential is diminished. We discuss, and provide some quantification of, these effects.
One of the main points stressed in this book is that there are bills that will be coming due for decades—including payments for disability and medical benefits. But how do we value these future costs? While all agree that a dollar in the future is worth less than a dollar today, the extent to which future costs are “discounted” is important (though changing the discount rate within a plausible range will not change the overall assessment, that this conflict will impose enormous economic costs, almost surely greater than every other war the United States has fought except World War II). In the final section, we explain the appropriate methodology for discounting.
Was the War About Oil?
LARRY LINDSEY, HEAD of Bush’s National Economic Council, claimed that “The successful prosecution of the war would be good for the economy” (cf. chapter 1). A key reason for this claim was the belief that it would keep oil prices low. As the Wall Street Journal editorial that same day argued, “the best way to keep oil prices in check is a short, successful war on Iraq.”1
This commonly accepted view was articulated clearly by Alan Greenspan, former chairman of the Federal Reserve. “If Saddam Hussein had been head of Iraq and there was no oil under those sands,” Greenspan said, “our response to him would not have been as strong as it was in the first Gulf War. And the second Gulf War is an extension of the first. My view is that Saddam, looking over his 30-year history, very clearly was giving evidence of moving towards controlling the Straits of Hormuz, where there are 17, 18, 19 million barrels a day” passing through.2 Greenspan noted in his memoirs that the fact that the war was “largely about oil” was “politically inconvenient.”3
There are other reasons that have caused many around the world to conclude that oil was the underlying motive for the war. When America went into Iraq, it went to great efforts to protect the oil assets, even as it failed to protect Iraq’s priceless antiquities, or (even more surprising from a military perspective) munitions supply. Moreover, while the weapons of mass destruction were put forward as the war’s rationale, there was at the time another country that was truly threatening to develop weapons of mass destruction: North Korea. But North Korea did not have oil and North Korea was not invaded. While America was focusing its attention on Iraq, North Korea became a nuclear power. Some interpreted the energy Bush put into getting debt relief for Iraq as motivated by oil: Iraq’s debt overhang cast a legal pallor over Iraq’s oil sales; creditors might go to courts to seize Iraqi oil in payment for what was owed. Only by ridding Iraq of this debt would it be possible for Iraq to sell its oil easily on global markets. The fact that Bush had long-standing ties to the oil industry and was knowledgeable about world oil markets, and that Iraq had one of the world’s largest reserves, made it plausible that oil was one of the factors on the president’s mind as he invaded Iraq.
From another perspective, however, the notion that the U.S. oil companies would be able to get Iraq’s oil for themselves was never very realistic. Some may have looked forward to a quick privatization of Iraqi oil, to be purchased on the cheap. But under the laws of occupation, that was not permitted (cf. chapter 6). Especially if there was more than a grain of truth in American promises about creating a democracy, there was little reason to believe that Iraqi politicians would simply execute America’s wishes. Oil is a global commodity, and they would have been under great pressure to get top dollar for their oil; American companies would have had to compete on an even footing with those from every other country. There is a limit to the number of regime changes that America could have engineered to get a government willing to execute its wish.
Moreover, there were other nations, such as Russia, claiming to have legal contracts that entitled them to develop some of Iraq’s oil resources. America could not simply assume that because it occupied Iraq, it could easily make these other claimants disappear. Indeed, when the Iraqi government, guided by U.S. legal advisers, cancelled a Russian contract, Russia retaliated by threatening to cancel its agreement to forgive $13 billion in Iraqi debt.4
In short, to the extent that oil did motivate the invasion, it was not based on a realistic analysis of the prospects of America gaining access for itself to an assured supply of oil. The belief that the United States invaded Iraq to get hold of its oil has, in fact, impeded reaching agreement on an oil law, viewed by many as critical for a future political settlement in the country. A response to the government’s draft oil law by 419 leading Iraqi academics, engineers, and oil industry experts stated, “it is clear that the government is trying to implement one of the demands of the American occupation,” and went on to argue that the law “lays the foundation for a fresh plundering of Iraq’s strategic wealth and its squandering by foreigners, backed by those coveting power in the regions, and by gangs of thieves and pillagers.”5
The Impact of the War on the Price of Oil
WHILE WE HAVE argued that the Iraq war’s disruption to the supply of oil is the single most important factor contributing to the soaring price, some analysts blame high global demand for oil, in particular from China. In this appendix, we explain why we believe the war is pivotal.
Before the Iraq war, China had had two decades of robust growth, and most analysts expected this to continue—with an accompanying increase in the demand for oil. And although global growth in 2003 and 2004 was stronger than many market analysts had anticipated, it was not markedly so. This can explain only some of the oil price rise. Moreover, well-functioning markets are not only supposed to anticipate changes in demand but to respond to changes in demand by increasing supply.6 Errors in one year are quickly corrected the next. It was anticipated that demand would be increasing in the coming years but that there would be a corresponding increase in supply, mostly from the Middle East—the low-cost supplier.
With oil this expensive, you would expect other oil-producing countries to start producing more. Many have (marginal) production costs far lower than current market prices.7 The anticipation of these supply-side responses would, in turn, drive down futures. The fact that there has not been this expected supply-side response, and that current and future oil prices are still so high, needs explaining. We think the Iraq war is a key part of the explanation.8
Had there been no war, and had the price of oil increased as a result of an unexpected increase in demand, the international community could have allowed Iraq to expand production, and this too could have brought down the price. Even if this had not happened, it is likely that production elsewhere, especially in the Middle East, would have increased. But the instability there has increased the risk of investing in that region; and because costs of extraction are so much lower in the Middle East, there has not been a commensurate supply response elsewhere. If stability is restored, prices will fall, and these investments elsewhere would turn a loss.9
Analyses of the Macroeconomic Impact of Higher Oil Prices
HERE, WE EXPLAIN why spending, say, $25 billion more on oil imports reduces GDP by a great deal—almost surely far more than the $37.5 billion we assume in our conservative scenario. That is, we explain why we think the oil multiplier (the ratio of the impact on GDP to the increased spending on oil imports) is greater than 1.5.10
The International Monetary Fund, for instance, has constructed econometric models that yield results with full effects (achieved over several years) that are almost four times as large as our estimate.11 Other studies suggest even larger multipliers.12
There are two possible explanations for the large discrepancies between the standard analyses, which often yield multipliers around 1.5, and these results. The first has to do with the analysis of global general equilibrium results. What gives rise to the multiplier is that money spent in the United States is spent again; as people buy goods and services, GDP is raised still more; and the higher GDP leads to still more expenditures, which in turn lead to still further increases in GDP. What limits the multiplier are leakages—money not spent “domestically” but taken out of the system, saved or spent abroad, or by government. In either case, the feedback of income into further expenditures stops. But if we take a global perspective, then the money spent abroad is part of the global economic system. Money spent, for example, on imports from Europe raises incomes in Europe, and some of that income is spent on imports from America. Thus, America still benefits. This would make the multiplier considerably larger.
Higher oil prices have depressed income in our major trading partners, Europe and Japan, and that has meant they have bought less from us than they otherwise would have, which in turn has increased the impact of higher oil prices on the U.S. economy.13 In Europe, inflationary pressures from higher energy prices most likely contributed to interest rates being higher than they otherwise would have been, especially given the European Central Bank’s single-minded focus on inflation. This has further weakened their economies—with knock-on effects on America’s.14 The European Union’s Stability and Growth Pact limits the ability of European governments to run deficits, which has meant they have not been able to respond adequately with fiscal policy; on the contrary, increased government spending on energy has meant there was less to spend on domestically produced goods and services, again contributing to the weakening of aggregate demand. In short, the direct effects of higher oil prices weakening Europe’s economy were made worse by these fiscal and monetary policy responses—enhancing the adverse effects on the U.S. economy.15
Second, standard analyses also focus only on short-run impacts—how higher oil prices today affect output today. But in this book, we are not concerned with these short-run impacts but with the total impact, year after year. When viewed from this long-run perspective, again, leakages are smaller. Money not spent this year (that is, savings) is spent in later years, stimulating income in those later years.16 The total impact of the oil price is accordingly much greater than the current impact (measured by the conventional multiplier).17
All these factors help explain why the “correct” multiplier, taking into account the full global effects, realized over many periods, may be a lot more than that generated by the models focusing only on the American economy in isolation (which generate multipliers of around 1.5), and why higher numbers such as those generated by the IMF model are reasonable.18 They also explain why we are confident that the multiplier we used in our moderate scenario is, in fact, highly conservative.
Was There Scope for Increasing Production?
WE HAVE ARGUED that had the United States not spent so much on oil and on the war in Iraq, our GDP would have been higher. The increased spending on American goods would have increased production. But that would have only been possible if production could have been increased. We explain here why we believe that throughout the Iraq war period, there was scope for increasing production—in some years by a considerable amount.
America has been operating below its potential. Potential output is defined as that output above which the rate of inflation starts to increase. In the late 1990s, America had an unemployment rate of 3.8%, and there did not appear to be any significant increases in inflation. In the Iraq war period, the unemployment rate has averaged more than 5%,19 suggesting that the economy could have expanded without inflationary pressures. It was lack of demand that was limiting output. There are two further pieces of evidence that support this view. First, the real unemployment rate—including disguised unemployment—has been high, markedly higher than, say, in 2000. Many Americans are working part time, involuntarily, because they cannot get full-time jobs. Many have dropped out of the labor market simply because they have found looking for work too discouraging—and are not included in the unemployment figures. And some have gone on disability because disability pays better than unemployment, and those who can get a doctor’s excuse do so.20
Second, pressure in the labor market is so weak that workers’ real wages (that is, taking into account inflation) have been falling relative to worker productivity—they are markedly below what they were at the beginning of the decade, or at the beginning of the war.21
If this analysis is correct, then there was ample scope for America to have expanded its output considerably—and certainly by the amount that it would have, had America not had to spend some $25 to $50 billion a year on imported oil, and had it switched some of the war spending to investments or other areas that would have stimulated the economy more.22
The Non-Quantifiable Macroeconomic Impacts of Higher Oil Prices
IN THE TEXT, we described the major quantifiable macroeconomic costs—arising from higher oil prices, switching government spending from productive investments to war expenditures, and increased deficits. We believe, however, that they represent an underestimation—perhaps a major underestimation—of the total costs to the economy. Here we examine this by looking at two categories of macroeconomic costs not considered in our earlier analysis.
First, the analysis of the cost of the higher oil price assumes that the only cost of the higher price is the increased transfer of dollars abroad to the oil exporters. It ignores adjustment costs and assumes that if the price increase is reversed, the damage is over. To put it another way, this simple model implies that if first the price goes up by $10 for one year, and then down by $20 for one year, and then is restored to its previous level, there is no cost. This is wrong. There is a cost to this volatility. The technology, for instance, that is best adapted to one set of prices will not be that appropriate for another. And these costs can be significant. This is consistent with macroeconomic studies which show large asymmetries between the impacts of increases and decreases in oil prices.23 Thus, this analysis of a seven-to eight-year period of high prices provides a significant underestimate of the true economic costs. We have not, however, provided an estimate of this additional cost.24
Second, most of our analysis focused on how the war—and the resulting oil price increases—dampened the American economy through demand-side effects. Because we were spending more money importing oil, and spending money in Iraq rather than at home, aggregate demand was lower. Earlier in this appendix, we argued that during most of the war period the economy could have produced more, if only there had been more demand.
Virtually all economists are agreed on two propositions. The first is that there is no such thing as a free lunch: While the Bush administration may have tried to persuade the American people that it could fight a war without any economic sacrifices, economists know otherwise. The second is that because Bush tried to fight the war without increasing taxes, the Iraq war has displaced private investment and/or government expenditures, including investments in infrastructure, R&D, and education: they are less than they would otherwise have been.25 The result is that the economy’s future potential and actual output over the long term will be lower, and in chapter 5, we have calculated by how much.26
Some economists, however, think that the supply-side effects—effects of the war on the economy’s production potential—are equally important even in the short run. If, for example, it were true that America’s economy was producing at its full potential, then those men and women from the National Guard and the Reservists sent to Iraq are not available for work in civilian jobs.27 These supply-side effects mount with the war: as the war continues, so do the casualties, producing increasing numbers who are partially or totally disabled and will never fully return to the labor force. Many of the returning veterans suffer from mental health conditions, which will interfere with their ability to be productive members of the labor force. We noted too in chapter 3 that many spouses and other family members have had to drop out of the labor force to care for returning disabled veterans, especially those needing medical care—in about one of five cases for a seriously injured veteran.28 We estimate that for the year 2006, the civilian labor supply has been reduced by approximately 140,000. Standard macroeconomic analyses suggest, at least in the short run, that GDP may fall (in percentage terms) by more than the value of the reduced employment. As the economy shrinks because of the lack of availability of labor, profit opportunities are also lost; and new bottlenecks appear. That is why the systemic cost may be so much greater than the direct costs of forgone labor.29
It is important to remember that the total number of servicemen and women involved in the Iraq conflict includes not just the 140,000–170,000 pairs of boots on the ground at any one time, but the far larger number who are between deployments, or based in military bases prior to being shipped to the theater. It also includes those providing logistical support. For the National Guard, we can argue that all of those mobilized are in effect part of the war effort, whether they are in active deployment or simply waiting to find out whether they will be required overseas.
Civilian GDP (GDP exclusive of what is being spent in Iraq) will be reduced too by the American contractors in Iraq. These are workers not available for producing consumer goods that individuals enjoy today or the investment goods that lead to future economic strength.30
Assuming that the loss in output is just proportional to the loss in labor implies for 2006 alone a loss of $13 billion, a total loss in GDP that is much larger than just the opportunity costs of these workers—the “microeconomic” costs discussed in chapter 4. Going forward, the losses to the labor force from those killed and disabled in the war will continue to increase, as will those who will have to drop out of the labor force to take care of them. It is likely that the number of Reservists and National Guard that are called upon will be further reduced; but, with prospects of large numbers permanently stationed in Iraq, the size of the military is likely to increase by some 92,000. This means that these supply-side losses to GDP are likely to continue, and even increase. Moreover, we should not really be focusing on the effect of the war on GDP—which values the bombs dropped in Iraq the same way as a newly built school or the salary of a research scientist making a breakthrough cure for some debilitating disease. We should really be looking at GDP net of resources spent on Iraq.
The most thorough analysis of the costs of the war using a comprehensive macroeconomic model was that of Allen Sinai, using the Sinai-Boston model with approximately 950 equations, and incorporating financial variables and their links to the “real” economy.31 He estimated that without the war (and ignoring the impact of the war on oil prices), real GDP growth would have been 0.2 percentage points higher, on average, over the period mid-2002 to mid-2005. The unemployment rate would have been 0.3 percentage points lower, on average, and almost 900,000 more non-farm payroll jobs per year would have been created. Assuming that impacts in future years are similar, the estimated macroeconomic effects are considerable—in excess of $200 billion.
Sinai calls attention to a further impact: there is a significant effect on government deficits. He calculates that the federal budget deficit would have been substantially lower. Tax receipts (personal and corporate, including capital gains, excise, and social insurance) would have been higher because of a better economy and a better stock market.
Determining the Discount Rate
IN THE TEXT, we argued that the appropriate discount rate should be 1.5%. That is the (real) rate at which government is able to borrow, which is why it is appropriate to use for purposes of evaluating impacts on the government budget. In chapters 4 and 5, however, we considered broader economic effects.
The debate about the appropriate discount rates has been contentious and confusing. There are two approaches. One focuses on how individuals trade off consumption (income) in different periods. The fact that individuals are willing to lend at 1.5% (real) interest means that this is their intertemporal trade-off, and so, in evaluating the impact on the well-being of individuals in society, this would seem to be the appropriate discount rate.
Other analysts argue that we should discount at the rate of the opportunity cost, the returns the funds might have generated had they been invested elsewhere. The calculations in the text take into account the opportunity cost of the funds; we analyze what GDP or national income might have been had the funds been spent, say, on investment, rather than on the war in Iraq. The question, though, is having analyzed the changes in output or consumption which might have been generated, how do we value an increase in consumption in the future relative to an increase in current consumption? The fact that individuals seem willing to trade off consumption today for that in the future using a 1.5% discount rate suggests this is the appropriate rate.
Three factors complicate this analysis. The first is that future consumption may not be enjoyed by the same individual but by future generations. How, in other words, should we evaluate at the margin consumption of the current generation versus that of future generations? A long philosophical tradition, dating back at least to the Cambridge economist Frank Ramsey in the 1920s, has argued that there is no justification for weighing future generations less than the current generation (except for a small factor, taking into account the risk of the extinction of the human race; and taking into account that because of productivity increases, future generations will be better off).32 In short, having calculated the changes in consumption that could have been generated, using a plausible value of opportunity cost (say, 6% to 8%), one then discounts these numbers back to present dollars using a low discount rate (say, 1% or 1.5%).33
Uncertainty presents the second complication. Some analysts discount at higher rates because the future is uncertain. This is an inappropriate, and potentially even dangerous, approach when it comes to valuing future uncertain costs. Discounting at a high rate (even 7%) means that we can effectively ignore such risks in the distant future. But if anything, the uncertainty should make us pay more, not less, attention to these risks. Our future health care and disability liabilities are examples of costs which, if anything, ought to be weighed more heavily because of the risks they represent: the uncertainty should, if anything, lead us to discount them at a lower rate. (Technically, the appropriate procedure entails converting costs and benefits into certainty equivalents, which increases costs and reduces benefits by the amount individuals would have been willing to pay to eliminate the risk, and discounting the certainty equivalents at the appropriate discount rate, say, 1.5%.) If uncertainty is increasing over time, this procedure entails increasing costs and decreasing benefits (relative to their average or mean values) over time. That is why, in the conservative approach taken here, when we evaluate the benefits that might have been generated by increasing investment had we not gone to war, we have looked at the consequences of using a higher discount rate, though in evaluating future costs (veterans’ health and disability), we have focused on the lower 1.5% rate.
The third complication is taxes on capital income. This introduces a discrepancy between individuals’ intertemporal trade-offs (how they value consumption today and in the future) and the return to capital (the opportunity cost). In evaluating the effects of the Iraq war, say financed by deficit, as we have noted, the appropriate procedure is to estimate what output would have been, and then (ignoring uncertainty and intergenerational effects) to discount the differences at individuals’ time preferences. Thus, if before-tax rates of return are 7%, and the marginal tax rate is approximately 40%, then the appropriate discount rate is approximately 4% (0.6 7%).34