The Myth of Military Myopia: Democracy, Small Wars, and Vietnam
Jonathan D. Caverley
International Security, Volume 34, Number 3, Winter 2009/10,
pp. 119-157 (Article)
Published by The MIT Press
For additional information about this article
http://muse.jhu.edu/journals/ins/summary/v034/34.3.caverley.html
Access Provided by University of Arizona at 08/11/11 4:37PM GMT
The Myth of Military Myopia
The Myth of Military
Myopia
Jonathan D. Caverley
Democracy, Small Wars, and Vietnam
P
owerful democracies
have often experienced frustration in ªghting small wars, particularly against
opponents employing unconventional strategies. The explanation for their
limited success is well known: wealthy, industrialized democracies tend to
pursue capital- and ªrepower-intensive strategies. The real puzzle is the problem’s venerability. Given the ample opportunities for learning, why then have
democracies not adopted a more effective means of counterinsurgency (COIN)
in small wars? If domestic constraints prevent democracies from spending
blood rather than treasure to ªght insurgencies, why ªght them in the ªrst
place?
In almost every investigation of this puzzle, the Vietnam War looms large as
a pivotal case.1 The U.S. failure in Vietnam, according to these accounts, epitomizes the inability of democratic militaries, in general, and the U.S. military, in
particular, to adopt an effective approach to overcoming an insurgency. The
Vietnam War is especially signiªcant for two generations of American military
intellectuals, who regard it as the paradigmatic case of organizational and
cultural inertia within the military. John Nagl describes the U.S. failure as “the
triumph of the institutional culture” in the U.S military, which counterproductively relied on “ªrepower and technological superiority.”2 Others use the
Jonathan D. Caverley is Assistant Professor of Political Science at Northwestern University.
The author would like to thank Ivan Arreguín-Toft, Alexander Downes, Charles Glaser, Jason
Lyall, John Mearsheimer, Edward Miller, John Nagl, Takayuki Nishi, Michael Noonan, Daryl
Press, Elizabeth Saunders, John Schuessler, Duncan Snidal, and seminar participants at the University of Chicago’s Program in International Security Policy, Dartmouth College’s International
Relations/Foreign Policy Working Group, and Northwestern University’s Buffett Center for International and Comparative Studies. He would also like to express his gratitude to the anonymous
reviewers, including Richard Betts, who waived anonymity. The Harry Frank Guggenheim Foundation and Harvard University’s Belfer Center for Science and International Affairs provided invaluable assistance for this research project.
1. Andrew Mack, “Why Big Nations Lose Small Wars: The Politics of Asymmetric Conºict,” World
Politics, Vol. 27, No. 2 (January 1975), pp. 175–200; Stephen Peter Rosen, “Vietnam and the American Theory of Limited War,” International Security, Vol. 7, No. 2 (Fall 1982), pp. 83–113; Gil Merom,
How Democracies Lose Small Wars: State, Society, and the Failures of France in Algeria, Israel in Lebanon,
and the United States in Vietnam (Cambridge: Cambridge University Press, 2003); and Ivan
Arreguín-Toft, “How the Weak Win Wars: A Theory of Asymmetric Conºict,” International Security, Vol. 26, No. 1 (Summer 2001), pp. 93–128.
2. John A. Nagl, Learning to Eat Soup with a Knife: Counterinsurgency Lessons from Malaya and Vietnam (Chicago: University of Chicago Press, 2005), p. 115; Andrew F. Krepinevich Jr., The Army and
Vietnam (Baltimore, Md.: Johns Hopkins University Press, 1986); and Larry E. Cable, Conºict of
Myths: The Development of American Counterinsurgency Doctrine and the Vietnam War (New York:
New York University Press, 1986). Historical accounts that blame military bureaucracy and culture
International Security, Vol. 34, No. 3 (Winter 2009/10), pp. 119–157
© 2010 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
119
International Security 34:3 120
Vietnam case to exemplify a more widespread apolitical and “machineminded” culture pervading all of American society.3 A less culturally oriented
explanation identiªes simple democratic cost aversion in Vietnam as the root
cause of U.S. failure to pursue an effective COIN strategy.4
The current unconventional conºicts in Iraq and Afghanistan have spurred
a renaissance of military-intellectual thought on both COIN and the myopia of
the U.S. military that prevents its successful application. Works in academia,
the military’s professional journals, and the popular press employ the Vietnam
War not simply for theory development but as an analogy for these two
conºicts.5 This approach raises more questions than it answers. Why has the
U.S. experience in Vietnam, sufªciently searing to have a “syndrome” attributed to it, failed to inform the subsequent conduct of counterinsurgency? If the
U.S. military is predisposed toward ineffective COIN, why does the civilian
leadership not step in or avoid such wars altogether? Why do voters ªxate on
reducing costs, while ignoring the reduced beneªts that result from such a
strategy? Why would democracies, supposedly the most prudent of regime
types, choose these risky wars and ªght them in such an unconstructive
manner?
This article answers these questions by arguing that to reduce the costs of
conºict for the relatively less wealthy voter, democratic leaders shift the burden of providing for the nation’s defense onto the rich by employing capital as
a substitute for military labor. Because the costs of ªghting an insurgency with
ªrepower are relatively low for the median voter compared to a more effective
but labor-intensive COIN approach, she will favor its use despite the diminished prospects of victory. This condition of moral hazard makes supporting a
capital-intensive military doctrine and small wars of choice rational policies
for the average voter.
In shedding light on why a democracy would enter a small war with the in-
for poor counterinsurgency doctrine but do not generalize to other small wars include Guenter
Lewy, America in Vietnam (New York: Oxford University Press, 1978); and Lewis Sorley, A Better
War: The Unexamined Victories and Final Tragedy of America’s Last Years in Vietnam (New York: Harcourt Brace, 1999).
3. Colin S. Gray, Irregular Enemies and the Essence of Strategy: Can the American Way of War Adapt?
(Carlisle, Pa.: Strategic Studies Institute, U.S. Army War College, March 1, 2006), p. 37.
4. Rosen, “Vietnam and the American Theory of Limited War.”
5. Gray, Irregular Enemies and the Essence of Strategy, p. 19; Douglas A. Ollivant, “Review of the
New U.S. Army/Marine Corps Counterinsurgency Field Manual,” Perspectives on Politics, Vol. 6,
No. 2 (June 2008), pp. 357–360; and Lt. Col. John A. Nagl, “A Better War in Iraq,” Armed Forces
Journal, August 2006, http://www.armedforcesjournal.com/2006/08/1931298/. For a more popular example, see Tom Bowman, “‘Clear and Hold’ Showing Results 40 Years Later,” Morning Edition, National Public Radio, October 13, 2008, http://www.npr.org/templates/story/story.php
?storyId!96350333.
The Myth of Military Myopia 121
tention of choosing a suboptimal strategy, this article challenges the dominant
explanation for ºawed counterinsurgency in the Vietnam War, which blames a
myopic military bureaucracy and culture for failing to adapt its conventional
doctrine to the conditions in Vietnam. Employing a variety of primary and secondary sources, the article reveals both military myopia’s limitations and its
need to be nested within a theory of civilian leaders and the public that elects
them. In doing so, the article marshals evidence that, contrary to much of the
conºict’s historiography, Lyndon Johnson’s administration played a crucial
role in rejecting a more labor-intensive COIN approach in favor of a capitalintensive strategy that it understood to be less effective but reºective of the
cost preferences of the average voter.
The article’s next section reviews previous attempts to explain democratic
counterinsurgency practices. It then lays out a competing theory of the median
voter and the cost distribution of conºict. The article then reviews the problematic counterinsurgency strategy that the United States pursued in Vietnam.
The subsequent section employs public opinion data to show that the public
preferences for the conduct of the war, as well as for the costs and beneªts, are
consistent with cost distribution theory. The next section examines the role of
civilian leaders in setting the U.S. strategy in Vietnam, particularly their efforts
to substitute treasure for blood through airpower, anti-inªltration barriers,
and a ªxation on the main force war. The concluding section suggests that the
theory has implications beyond the case of Vietnam.
How to Lose a Small War
The venerable U.S. Marine Corps’ Small Wars Manual deªnes a small war as
one “undertaken under executive authority, wherein military force is combined with diplomatic pressure in the internal or external affairs of another
state whose government is unstable, inadequate, or unsatisfactory for the preservation of life and of such interests as are determined by the foreign policy of
our Nation.”6 As the ªrst phrase implies, small wars do not require the mobilization of the country, although this does not obviate the need for public support. They are fought by a powerful state against a weaker state or nonstate
actor (“weak actor” for simplicity). A small war is one of choice; it may be consistent with the strong state’s grand strategy but not essential to it. The strong
state’s aims are limited or political, and success often requires the weak actor’s
compliance.
6. United States Marine Corps, Small Wars Manual (Washington, D.C.: U.S. Government Printing
Ofªce, 1940), p. 1, http://www.smallwars.quantico.usmc.mil/SWM/1215.pdf.
International Security 34:3 122
Because strong states tend to enjoy overwhelming conventional military superiority, weak actors will likely resort to unconventional strategies such as insurgency or terrorism.7 Fighting an unconventional war is a daunting task
even for powerful states. Usually it demands tremendous investments in intelligence gathering and a deep understanding of a foreign culture. Success requires gaining the allegiance, or at least acquiescence, of local noncombatants
by providing personal security and economic stability. Firepower, when not
used with the utmost discrimination, will likely have counterproductive effects. In general, no substitute exists for boots on the ground; the ratio of personnel to population required for nation building has stayed roughly stable at
20 per 1,000 since the end of World War II.8
These principles of a successful COIN strategy have remained largely consistent over at least the past half century.9 Indeed, a remarkable amount of
agreement exists on how states lose small wars. Ivan Arreguín-Toft demonstrates how a conventional offensive campaign against a guerrilla warfare
strategy will likely result in a win (or at least a “non-loss”) for the guerrillas.10
Jason Lyall and Isaiah Wilson systematically test and ªnd support for the
proposition that mechanized militaries are less effective in small wars because
of attendant collateral damage, a poor intelligence-gathering ability, and an inability to secure the population.11
why focus on democracies?
Democracies do not appear as successful at ªghting insurgencies relative to
their track record in conventional wars. The mosaic plots in ªgures 1 and 2 use
two data sets to compare the performance of democracies and nondemocracies
in major wars (battle deaths exceeding 1,000) to outcomes of conºicts where a
7. Arreguín-Toft, “How the Weak Win Wars.”
8. James T. Quinlivan, “Burden of Victory: The Painful Arithmetic of Stability Operations,” RAND
Review, Vol. 27, No. 2 (Summer 2003), p. 28.
9. Classic works on COIN include David Galula, Counterinsurgency Warfare: Theory and Practice
(Westport, Conn.: Praeger, 2006); and Robert Thompson, Defeating Communist Insurgency: The Lessons of Malaya and Vietnam (St. Petersburg, Fla.: Hailer, 1966). More recent versions with similar recommendations include David Kilcullen, The Accidental Guerrilla: Fighting Small Wars in the Midst of
a Big One (New York: Oxford University Press, 2009); and Kalev I. Sepp, “Best Practices in Counterinsurgency,” Military Review, Vol. 85, No. 3 (May/June 2005), pp. 8–12. For the remainder of
the article, I use “COIN” to denote this effective, labor-intensive form of stopping an insurgency;
“counterinsurgency” refers to any strategy used against insurgents including the ªrepowerintensive U.S. version in Vietnam.
10. Arreguín-Toft, “How the Weak Win Wars.”
11. Jason Lyall and Isaiah Wilson III, “Rage against the Machines: Explaining Outcomes in Counterinsurgency Wars,” International Organization, Vol. 63, No. 1 (Winter 2009), pp. 67–106. For a remarkable demonstration of bombing’s ineffectiveness in the Vietnam War counterinsurgency, see
Matthew Adam Kocher, Thomas B. Pepinsky, and Stathis N. Kalyvas, “Aerial Bombardment, Indiscriminate Violence, and Territorial Control in Insurgencies: Evidence from Vietnam,” Working
Paper (New Haven, Conn.: Yale University Press, 2009).
The Myth of Military Myopia 123
Figure 1. Comparison of Outcomes of Wars (more than 1,000 battle deaths) by Regime,
1816–1990
SOURCE: Alexander B. Downes, “How Smart and Tough Are Democracies? Reassessing The-
ories of Democratic Victory in War,” International Security, Vol. 33, No. 4 (Spring 2009),
pp. 9–51.
state ªghts an insurgency outside of the state’s territory. Whereas democracies
win 62 percent of the larger, generally conventional wars, they win only
47 percent of the counterinsurgencies (nondemocracies win 40 and 58 percent,
respectively).12 Figure 2 shows that democracies are no more likely than
nondemocracies to win against an insurgency, and are considerably more
likely to ªght to a draw.13
While arguing that regime type does not affect overall performance against
insurgencies, Lyall notes that democratic counterinsurgency efforts are more
12. Figure 1’s data are from a source skeptical of democracies’ performance in these wars. Alexander B. Downes, “How Smart and Tough Are Democracies? Reassessing Theories of Democratic
Victory in War,” International Security, Vol. 33, No. 4 (Spring 2009), pp. 9–51. Other sources claim
that democracies win as much as 93 percent of such conºicts. Dan Reiter and Allan C. Stam, Democracies at War (Princeton, N.J.: Princeton University Press, 2002).
13. A multinomial logit analysis of “democracy” on war outcome shows that democracies are just
as likely to lose to an insurgency as nondemocracies and signiªcantly more likely to ªght to a
draw (with a coefªcient for democracy of 1.872 and a standard error of 0.85). Data from Lyall and
Wilson, “Rage against the Machines.”
International Security 34:3 124
Figure 2. Comparison of Outcomes of Third-Party Interventions against Insurgencies by
Regime, 1808–2002
SOURCE: Jason Lyall and Isaiah Wilson III, “Rage against the Machines: Explaining Outcomes
in Counterinsurgency Wars,” International Organization, Vol. 63, No. 1 (Winter 2009),
pp. 67–106.
likely to be wars of choice abroad and tend to employ heavily capitalized militaries.14 Once these and other factors are controlled for, democracy has little independent effect on war outcome. But if democracies are more likely to select
challenging third-party conºicts, and are more likely to use a capital-intensive
doctrine while doing so, then regime type may well play a role. Even if regime
type makes little difference, this in itself is a puzzle given the large body of research claiming that democracies pick unfair ªghts that they tend to win.
the puzzle for democratic exceptionalism
Democracies’ poor track record in small wars challenges the liberal consensus
that democratic states tend to pursue exceptionally moderate and successful
14. Jason Lyall, “Do Democracies Make Inferior Counterinsurgents? Reassessing Democracy’s Impact on War Outcomes and Duration,” International Organization, forthcoming.
The Myth of Military Myopia 125
foreign policies.15 Many of this research program’s ªndings rest on the assumption that in democracies the costs of war are internalized; all costs and
beneªts of a decision are accounted for by the actor responsible for setting policy. Fred Chernoff describes the difference between democracies and other
kinds of regimes in this regard: “Citizens and subjects—rather than presidents
and monarchs—ªght in wars, die in wars, and pay taxes to ªnance wars.
In most cases, it is not in the citizen’s self-interest for the state to go to
war.”16 Conversely, shielding the decisionmaker from the costs of war can
lead to aggressive behavior. The most comprehensive statement of this costinternalization mechanism suggests that democratic leaders respond to the
voters’ cost-beneªt calculation by providing public goods, including security
and military victory, both efªciently and in abundance.17
Although the logic of exceptionalism suggests that democracies pursue superior foreign policies, many studies focusing on small wars claim that democratic cost aversion leads to ºawed warªghting.18 Stephen Rosen argues that
15. Downes, “How Smart and Tough Are Democracies?” lays out a general empirical critique of
this program, including an examination of the Vietnam War as a deviant case.
16. Fred Chernoff, “The Study of Democratic Peace and Progress in International Relations,” International Studies Review, Vol. 6, No. 1 (Spring 2004), p. 54. See also Reiter and Stam, Democracies at
War, p. 121; and Randolph M. Siverson, “Democracies and War Participation: In Defense of the Institutional Constraints Argument,” European Journal of International Relations, Vol. 1, No. 4 (December 1995), p. 483. This mechanism is used to explain why democracies, ªrst, ªght shorter wars—
see D. Scott Bennett and Allan C. Stam III, “The Declining Advantages of Democracy: A Combined
Model of War Outcomes and Duration,” Journal of Conºict Resolution, Vol. 42, No. 3 (June 1998),
pp. 344–366; and Branislav L. Slantchev, “How Initiators End Their Wars: The Duration of Warfare
and the Terms of Peace,” American Journal of Political Science, Vol. 48, No. 4 (October 2004), pp. 813–
829—second, prefer to negotiate—see Bruce Bueno de Mesquita, Alastair Smith, Randolph M.
Siverson, and James D. Morrow, The Logic of Political Survival (Cambridge, Mass.: MIT Press, 2003);
and Darren Filson and Suzanne Werner, “Bargaining and Fighting: The Impact of Regime Type on
War Onset, Duration, and Outcomes,” American Journal of Political Science, Vol. 48, No. 2 (April
2004), pp. 296–313—third, win the wars they do initiate—see Bueno de Mesquita et al., The Logic of
Political Survival, chap. 6; Reiter and Stam, Democracies at War; and Siverson, “Democracies and
War Participation”—and, fourth, spend less money on defense in peacetime but devote more to
the effort in wartime—see Benjamin O. Fordham and Thomas C. Walker, “Kantian Liberalism, Regime Type, and Military Resource Allocation: Do Democracies Spend Less?” International Studies
Quarterly, Vol. 49, No. 1 (March 2005), pp. 141–157; and Benjamin E. Goldsmith, “Defense Effort
and Institutional Theories of Democratic Peace and Victory: Why Try Harder?” Security Studies,
Vol. 16, No. 2 (April–June 2007), pp. 189–222.
17. Bruce Bueno de Mesquita and George W. Downs, “Intervention and Democracy,” International
Organization, Vol. 60, No. 3 (Summer 2006), pp. 627–649; and Bueno de Mesquita et al., The Logic of
Political Survival. See also David A. Lake, “Powerful Paciªsts: Democratic States and War,” American Political Science Review, Vol. 86, No. 1 (March 1992), pp. 24–37.
18. Mack, “Why Big Nations Lose Small Wars”; Zeev Maoz, “Power, Capabilities, and Paradoxical
Conºict Outcomes,” World Politics, Vol. 41, No. 2 (January 1989), pp. 239–266; John E. Mueller,
“The Search for the ‘Breaking Point’ in Vietnam: The Statistics in a Deadly Quarrel,” International
Studies Quarterly, Vol. 24, No. 4 (December 1980), pp. 497–519; and Patricia L. Sullivan, “War Aims
and War Outcomes: Why Powerful States Lose Limited Wars,” Journal of Conºict Resolution, Vol. 51,
No. 3 (June 2007), pp. 496–524.
International Security 34:3 126
the Johnson administration chose a futile signaling strategy in the Vietnam
War to minimize losses. Gil Merom identiªes a catch-22: democracies build
ªrepower-intensive, low-manpower militaries to reduce “the number and/or
exposure to risks of soldiers,” but consequently they must rely on “higher and
less discriminating levels of violence,” a policy that leads to criticism from the
“educated middle class.” This cost aversion results in a “post-heroic warfare”
employed by “Western democracies conducting non-existential wars in which
their readiness to sacriªce is relatively low.”19
These explanations clarify why democracies may prefer a military doctrine
poorly suited for small wars, but they cannot account for their insistence on
ªghting them anyway. Surely a pragmatic state would rather not ªght a war at
all than ªght one it is likely to lose. Patricia Sullivan identiªes the problem:
“Extant theories cannot explain why militarily preponderant states regularly
make poor strategic choices,” but Sullivan’s argument that the aims associated
with small wars can lead to increased uncertainty over the likely costs does not
ªll this void; a strategic actor should recognize this and adjust for the larger
downside risk before entering a conºict.20
military myopia and other sources of nonstrategic behavior
Organizational and cultural theories about the military’s role in strategy
development claim to ªll this explanatory breach. These theories argue that
without sufªcient pressure from political leaders, elements of the national security structure, particularly the military, will pursue their own ends with little
regard for grand strategy. Barry Posen and Jack Snyder focus on bureaucratic
forces pushing militaries toward adopting offensive doctrines, whereas
Elizabeth Kier argues that military culture is of greater importance and not
simply limited to a preference for the offense.21 A focus on military culture is
prominent in work addressing U.S. conduct of small wars; Eliot Cohen, for
example, states that “the most substantial constraints on America’s ability to
conduct small wars result from the resistance of the American defense estab-
19. Rosen, “Vietnam and the American Theory of Limited War”; Merom, How Democracies Lose
Small Wars, pp. 21–22; Avi Kober, “The Israel Defense Forces in the Second Lebanon War: Why the
Poor Performance?” Journal of Strategic Studies, Vol. 31, No. 1 (February 2008), pp. 3–40; and Edward N. Luttwak, “A Post-Heroic Military Policy,” Foreign Affairs, Vol. 75, No. 4 (July/August
1996), pp. 33–44.
20. Sullivan, “War Aims and War Outcomes,” p. 497.
21. Barry R. Posen, The Sources of Military Doctrine: France, Britain, and Germany between the World
Wars (Ithaca, N.Y.: Cornell University Press, 1984); Jack L. Snyder, The Ideology of the Offensive: Military Decision Making and the Disasters of 1914 (Ithaca, N.Y.: Cornell University Press, 1984); and Elizabeth Kier, Imagining War: French and British Military Doctrine between the Wars (Princeton, N.J.:
Princeton University Press, 1997).
The Myth of Military Myopia 127
lishment to the very notion of engaging in such conºicts, and from the unsuitability of that establishment for ªghting such wars.”22
These approaches support the argument of Robert Komer, President
Johnson’s principal counterinsurgency adviser, that allowing the military to
“do its thing” during wartime is a mistake.23 The Vietnam War is often described as the poster child for this pathology. Military myopia arguments
claim that the U.S. military, particularly the Military Assistance Command
Vietnam (MACV) commander, Gen. William Westmoreland, displayed “an utter obliviousness to the political nature of the war” by rejecting an alternative
COIN or paciªcation strategy that was simultaneously more effective and less
casualty intensive.24
Whereas military myopia requires a civil-military disconnect, an alternative
suggests that the U.S. military simply reºects the cultural preferences of the
people it serves. Colin Gray describes the United States’ “public, strategic, and
military culture” as “not friendly to the means and methods necessary for the
waging of warfare against irregular enemies.”25 Americans, Gray claims, are
profoundly apolitical when it comes to war.26 Apart from its tautological nature, a theory based on the premise that “America is what it is” cannot explain
why the difªculty that democracies face in ªghting small wars is not isolated
to the United States.27
A Theory of Redistribution and Military Doctrine
This section offers a theory of how a rational actor, the average voter in a democracy, can favor what appears to be a nonstrategic policy. To do so, it uses
the core logic of a research program claiming that this sort of behavior should
rarely happen in democracies. The theory presented here shares three important assumptions with the cost-internalization logic of democratic exceptionalism: the distribution of costs within the state affects its pursuit of security; a
democratic government’s provision of security is a public good; and voters
22. Eliot A. Cohen, “Constraints on America’s Conduct of Small Wars,” International Security, Vol.
9, No. 2 (Fall 1984), p. 165.
23. Robert W. Komer, Bureaucracy Does Its Thing: Institutional Constraints on U.S.-GVN Performance
in Vietnam (Santa Monica, Calif.: RAND, 1973).
24. Jeffrey Record, Beating Goliath: Why Insurgencies Win (Washington, D.C.: Potomac, 2007), p. 121;
and Krepinevich, The Army and Vietnam, p. 21. Krepinevich challenges what many consider the
U.S. Army’s received wisdom embodied by Harry G. Summers Jr., On Strategy: A Critical Analysis
of the Vietnam War (Mineola, N.Y.: Dover, 2007). This article agrees with Summers that civilians set
much of the ground strategy, but it disagrees that these civilians pushed the army toward COIN.
25. Gray, Irregular Enemies and the Essence of Strategy, p. vi.
26. Ibid., p. 30.
27. Ibid., p. 7; and Record, Beating Goliath.
International Security 34:3 128
“take a reasonably level-headed cost-beneªt approach in forming attitudes towards military missions.”28 I relax the claim that costs are always internalized
within democracies, however, arguing that the average voter’s share may be
much lower than the state’s per capita costs.
Even in democracies, wealth is not distributed equally within any given
state; the person with median income is less well off than someone possessing
the mean. A political-economic approach, the Meltzer-Richard hypothesis,
suggests that if the median voter can set a tax rate and spend the revenue on a
service available to all citizens, she will take advantage of the potential for redistribution.29 Even with a ºat tax on income, the wealthy will pay a larger
portion of the costs for a public good enjoyed by all. For example, in 2005 the
ªfth of the population with the highest incomes paid 69 percent of all U.S. federal tax revenue, and the middle ªfth paid only 9 percent.30 Similarly, the median voter will prefer to tax capital more heavily than labor, because labor
income is distributed more equally than capital income.31
How these taxes are spent plays a most important role in establishing the
redistributive nature of the public good of defense. Military doctrine, the
means by which military power is developed and exercised, can be stylized as
a production function consisting of two factors—capital (e.g., tanks, planes,
ammunition, and even training) and labor (soldiers, sailors, etc.)—as well as
the technology that allows one factor of production to serve as a substitute for
another.32 Capital and labor are imperfect replacements and show diminishing
returns; given a hundred tanks and ten soldiers, adding another tank will not
produce as much capability as adding another soldier. The type of conºict affects substitutions as well; it is much harder to substitute capital for labor
when ªghting an unconventional opponent.
28. On the last assumption, see Christopher Gelpi, Peter D. Feaver, and Jason Reiºer, Paying the
Human Costs of War: American Public Opinion and Casualties in Military Conºicts (Princeton, N.J.:
Princeton University Press, 2009), p. 2.
29. Allan H. Meltzer and Scott F. Richard, “A Rational Theory of the Size of Government,” Journal
of Political Economy, Vol. 89, No. 5 (October 1981), pp. 914–927; Alberto Alesina and Dani Rodrik,
“Distributive Politics and Economic Growth,” Quarterly Journal of Economics, Vol. 109, No. 2 (May
1994), pp. 465–490; and Torsten Persson and Guido Tabellini, Political Economics: Explaining Economic Policy (Cambridge, Mass.: MIT Press, 2000). For a recent use of median voter theory and redistribution, see Daron Acemoglu and James A. Robinson, Economic Origins of Dictatorship and
Democracy (New York: Cambridge University Press, 2005).
30. Edward Harris, Historical Effective Federal Tax Rates: 1979 to 2005 (Washington, D.C.: Congressional Budget Ofªce, December 2007), http://www.cbo.gov/doc.cfm?index!8885.
31. Persson and Tabellini, Political Economics, pp. 117–122.
32. This article does not distinguish among strategy, operations, and tactics; substitution of capital
for labor occurs at all levels. I have settled on the term “doctrine,” deªned by the U.S. Defense Department as “fundamental principles by which the military forces or elements thereof guide their
actions in support of national objectives,” as the appropriate catch-all phrase. See http://
www.dtic.mil/doctrine/jel/doddict/data/d/01744.html.
The Myth of Military Myopia 129
Tax revenue can pay for both the capital and labor inputs. Personnel also can
be supplied through conscription, a tax on a citizen’s labor rather than income.
Even if the odds of being conscripted are equally distributed, the median voter
will demand that a larger amount of the military budget go toward the purchase of capital to reduce the risk of conscription. In cases where existing
threats do not currently justify resorting to conscription, military capitalization
will still to a large degree determine a draft’s future likelihood. The median
voter normally will be happy with an expensive, all-volunteer military; but
once the level of threat creates a demand for labor that reaches into the middle
class, the voter will support a conscripted military where draftees are protected by large amounts of capital.33
Casualties are also a public bad: no one wants to see their fellow citizens die.
The less wealthy are more likely to be drafted and to join an all-volunteer
force; they may gain jobs from domestic weapons manufacturing; and they often regard military service as a means of acquiring human capital. Conscription is therefore an important, but not the only, reason why militaries with
large amounts of labor can be a public bad. The median voter will accept a
higher tax, what the British socialist Sidney Webb called the “conscription of
riches,” to build highly capitalized militaries both in peace and in war, because
such militaries redistribute money and skills through jobs and training as well
as reduce the risk of conscription and casualties.
The median voter theory outlined above does not claim to perfectly capture
how policy is made in a democracy, nor does it argue that one’s relative income determines one’s position on foreign policy. Rather, this simple theory
suggests an equally simple insight: a military doctrine that privileges capital
over labor will reduce the costs of conºict for an important swath of voters. A
capitalized military not only results in many voters doing less of the ªghting
themselves, but also allows someone else’s resources to fund the costs of war.
Politicians should respond accordingly. This distribution of costs explains how
a state’s seemingly nonstrategic behavior may be in the interests of important
rational actors within a democracy.
Because of its redistributive nature, a capitalized military doctrine can lead
to moral hazard, which arises when perverse incentives encourage actors to
pursue risky behavior. For example, drivers with auto theft coverage are more
likely to park on the street than pay for secure parking. Many domestic government programs merge moral hazard with the Meltzer-Richard effect dis33. Joseph Paul Vasquez III, “Shouldering the Soldiering: Democracy, Conscription, and Military
Casualties,” Journal of Conºict Resolution, Vol. 49, No. 6 (December 2005), pp. 849–873. If conscription falls more heavily on the less wealthy, then capital will be further favored.
International Security 34:3 130
cussed above. Deposit insurance uses government backing to insure bank
deposits up to a certain limit, a redistributive public good. Because the insurance applies regardless of the bank (subject to government regulations), an
individual has little motivation to consider the bank’s solvency. Indeed,
she is likely to choose the higher interest provided by a bank making risky
investments.
Regarding defense provision, a lack of cost internalization creates an incentive for the median voter to support risky behavior: that is, using a capitalintensive military in conºicts where substitutability is low because the
decreased likelihood of winning is outweighed by the lower costs of ªghting
in such a manner. If the median voter’s risky behavior is in effect being subsidized by the wealthy, democratic leaders sensitive to this voter’s costs will
pursue strategies that make success less likely. I argue that this is what
happened in Vietnam.34
Why Did the United States Fight Poorly in Vietnam?
This article examines the U.S. military’s counterinsurgency practices in
Vietnam during the Johnson administration, when nearly all major escalation
and warªghting decisions were made. It also brieºy visits Richard Nixon’s administration to illustrate the essential continuity of U.S. military strategy in
Vietnam. Consensus exists on the reasons why the choice of strategies contributed to failure. First, the United States pursued a skewed balance of effort between ªghting the enemy’s main and insurgent forces. Although throughout
the war the United States had to face both an insurgency and a conventional,
“main force” threat, it focused disproportionately on ªnding, ªxing, and destroying conventional enemy units. “Paciªcation”—that is, securing noncombatants from predation by Vietcong (VC) guerrillas and ensuring some
economic stability—was arguably the more important task from the major U.S.
escalations of 1966 onward and yet was given short shrift.
Second, the United States employed the few resources it devoted to combating the insurgency ineffectively. Rather than rooting out guerrilla elements
from populated areas, establishing secure spaces for South Vietnamese civilians, and engaging in civil development, the United States sought to use
ªrepower to interdict supplies for the insurgency, engage in strategic bombing
34. Many factors besides redistributive preferences—labor productivity, wealth, population age,
level of education, and geography—both inºuence military doctrine and correlate with democracy. The United States and its allies built high-quality militaries to counteract the Soviet Union’s
quantitative advantage. By concentrating on warªghting across a long conºict, I essentially hold
these other factors constant.
The Myth of Military Myopia 131
to make North Vietnam pay costs for its support, and pursue search-anddestroy missions to kill enemy personnel at a rate exceeding the reinforcement
rate. U.S. forces were for the most part excluded from paciªcation efforts, leaving these tasks to a South Vietnamese military and government clearly not
competent in these missions. This article explains why airpower, physical barriers, and a main force–oriented attrition strategy were employed by the
United States against Vietnamese guerrillas (rather than as a necessary component of combating enemy conventional forces), as well as why the counterinsurgency received less attention relative to the main force war. In doing so, this
article tests its theory against its rivals by presenting evidence that cost distribution theory explains more aspects of the war than do other theories. This approach also helps to illustrate the theory’s causal chain, inspiring conªdence
that the cost distribution mechanism was a necessary cause for the Johnson administration’s poor military strategy in the Vietnam War.
A successful test of cost distribution theory’s causal logic requires empirical
support for three propositions. First, the public must back a capital-intensive
approach to limited war. This does not preclude support for sending soldiers
into harm’s way, nor does cost distribution theory require comprehensive
thinking on counterinsurgency doctrine by the electorate. Rather, the theory
suggests that broad sections of the public will assess the costs of the conºict in
blood and treasure and will favor the latter. Second, government ofªcials must
acknowledge the public (not necessarily an explicitly named “median voter”)
as the source of pressure to ªght a capital-intensive campaign. Unlike the electorate, policymakers should understand the hazards of applying such an approach to counterinsurgency and recommend it anyway. Third, government
ofªcials must explicitly direct the military to ªght accordingly.
Evidence that the Johnson administration, important members of Congress,
the U.S. military, and even the American public shared a realistic assessment of
the limited nature, stakes, and prospects for success in Vietnam supports cost
distribution theory over other explanations. Showing that the administration
understood the superiority of a COIN approach and rejected it on domestic
political grounds also undermines the strategic culture arguments. Evidence
that General Westmoreland and other military experts recommended a more
COIN-oriented strategy only to see this recommendation rejected by President
Johnson and his advisers would undermine the military myopia case in favor
of cost distribution. Finally, unlike the alternatives, cost distribution theory
suggests that the choice to pursue counterinsurgency with conventional forces
and tactics should be consistent, despite feedback that the approach is not
working. Democratic exceptionalism, military myopia, and even strategic culture explanations suggest that a democracy eventually adjusts in the face of
International Security 34:3 132
harsh wartime lessons. Cost distribution theory argues that, short of reducing
the average voter’s inºuence, a democracy is unlikely to “learn” to conduct effective COIN.35
Public Opinion on the Vietnam War
Polling data show that a large portion of the American public remained
vaguely hawkish, if poorly informed, about the Vietnam War, through the Tet
offensive of 1968.36 Afterward, even though a ªve-to-three majority of the public viewed the decision to go to war as a mistake, the same ratio wanted to win
the war by escalating U.S. efforts.37 Once this hawkish consensus began to
break down in 1968, the role of income in public opinion of the war grew
signiªcant, with poorer people being more likely to support the war.38 A closer
look at the polling data reveals a public relatively realistic about the prospects
for limited success in Vietnam and the means they were willing to employ to
achieve it.
public recognition and support for limited war
Critics describe U.S. strategic culture as suffering from an “apolitical view of
war, which encourages the pursuit of military victory for its own sake”; polling data, however, do not support this claim.39 At the time of the 1965 U.S. escalation of the Vietnam War, 64 percent of Gallup poll respondents supported
greater involvement, yet only 29 percent thought that victory was likely. Another 30 percent predicted a stalemate.40 Across several identical polls from
January 1966 through the end of 1972, large majorities (ranging from 53 to
77 percent) agreed that the war was likely to end in a “compromise peace
settlement.”41
35. Within space constraints, this article includes representative citations on war strategy and domestic politics from every major adviser to Lyndon Johnson from 1964 to 1968: the national security advisers, secretaries of state and defense, chair of the Joint Chiefs of Staff, personal advisers
such as Maxwell Taylor and Robert Komer, as well as the most pivotal deputies: William Bundy,
Nicholas Katzenbach, and John McNaughton. Vice President Hubert Humphrey wrote a prescient
analysis of the insurgency and was subsequently marginalized by Johnson. William C. Gibbons,
The U.S. Government and the Vietnam War: Executive and Legislative Roles and Relationships, Part 4:
July 1965–January 1968 (Princeton, N.J.: Princeton University Press, 1995), p. 20.
36. John E. Mueller, War, Presidents, and Public Opinion (New York: Wiley, 1987), p. 54.
37. Leslie H. Gelb and Richard K. Betts, The Irony of Vietnam: The System Worked (Washington, D.C.:
Brookings Institution Press, 1979), p. 172.
38. Scott Sigmund Gartner and Gary M. Segura, “Race, Casualties, and Opinion in the Vietnam
War,” Journal of Politics, Vol. 62, No. 1 (February 2000), pp. 115–146.
39. Record, Beating Goliath, p. x.
40. Gelb and Betts, The Irony of Vietnam, pp. 129–130.
41. Survey by Gallup Organization, January 1–5, May 19–24, 1966; May 11–16, November 16–21,
1967; February 2–6, March 2–7, 1968; March 1971; and November 10–13, 1972 (hereafter “Gallup”),
iPOLL Databank, Roper Center for Public Opinion Research, University of Connecticut.
The Myth of Military Myopia 133
In addition to maintaining a realistic assessment of the outcome, the largest
group in most surveys supported the pursuit of limited U.S. aims in Vietnam.
A Harris poll in December 1964 found that 40 percent of the American public
approved of “continuing support for the anti-Communist government,” and
roughly equal portions supported withdrawal or “bombing North Vietnam.”42
In a November 1966 poll, only 7 percent preferred to pull troops out, and another 5 percent supported a “neutralist South Vietnam.” Fifty-seven percent
supported both sides withdrawing “under the United Nations,” and 31 percent advocated the pursuit of “total military victory.” In February 1967 the
same question gained similarly meager levels of support for the ªrst two options (6 and 7 percent, respectively), whereas 44 percent supported the UN
option, and a striking 43 percent supported total victory. The same question in
May 1967 produced nearly identical results.43
public preference for capital over labor
Although many poll respondents were unsure about the strategy and tactics
being pursued by the U.S. military in Vietnam, those expressing an opinion appeared realistic in assessing their efªcacy. Public opinion was hawkish on the
issue of bombing, far less so on the role of ground forces.
Figures 3a to 3d present poll data from June 1965, taken at the cusp of accelerated U.S. involvement in Vietnam. About 80 percent of respondents were either unsure or skeptical that airpower alone could stop North Vietnamese
inªltration of the South (ªgure 3a), yet a large majority (58 percent) favored
continued bombing of the North in retaliation (3b). The results are far more
mixed when ground operations are covered. The percentage favoring a troop
increase is also large (62 percent, ªgure 3c), but only under limited circumstances; respondents were told that the troops would be defending the South
“this summer” against a communist “land offensive to drive the Americans
out of South Vietnam.” Figure 3d shows that only 23 percent favored
“carry[ing] the ground war into North Vietnam”; a plurality (46 percent)
wanted the U.S. military to “hold the line”; and more respondents preferred
negotiations over escalation.44
A poll taken six months later (ªgure 4a) shows that 52 percent of respondents supported the Johnson administration’s decision to end the second
“Christmas” bombing pause (22 percent opposed). While a smaller plurality
42. John D. Stempel, “Policy/Decision Making in the Department of State: The Vietnamese Problem, 1961–1965,” Ph.D. dissertation, University of California, Berkeley, 1966, pp. 249–250.
43. Mueller, War, Presidents, and Public Opinion, p. 87.
44. Survey by Louis Harris and Associates, “Harris 1966 Survey, Nos. 1522 and 1531,” in Harris
Collection, Odum Institute, Louis Harris Data Center, University of North Carolina, http://
arc.irss.unc.edu/dvn/dv/odvn/faces/StudyListingPage.xhtml?mode!1&collectionId!4.
(d) Carry ground war into North, negotiate, or hold the line to prevent
Communists taking over South Vietnam?
(b) Favor continued bombings in North in retaliation for Communist raids in
South?
North Carolina, http://arc.irss.unc.edu/dvn/dv/odvn/faces/StudyListingPage.xhtml?mode=1&collectionId=4.
SOURCE: Survey by Louis Harris and Associates, “Harris 1966 Survey, No. 1522,” in Harris Collection, Odum Institute, Louis Harris Data Center, University of
(c) Communists have said they will mount a land offensive to drive the
Americans out of South Vietnam. Favor the U.S. sending in more troops
to defend South Vietnam this summer?
(a) Can infiltration of South Vietnam be stopped by air raids on North or
only by sending more U.S. troops to fight them?
Figure 3. American Public Opinion on U.S. Bombing and Ground Forces, June 1965
The Myth of Military Myopia 135
Figure 4. American Public Opinion on U.S. Bombing Pause and Enclave Strategies, June
1966
(a) Administration more right or wrong?
Not continuing bombing pause.
(b) Administration more right or wrong?
Not limiting U.S. ground action to a few
strong coastal areas (enclaves).
SOURCE: Survey by Louis Harris and Associates, “Harris 1966 Survey, No. 1623,” in Harris
Collection, Odum Institute, Louis Harris Data Center, University of North Carolina, http://
arc.irss.unc.edu/dvn/dv/odvn/faces/StudyListingPage.xhtml?mode=1&collectionId=4.
supported doing more than “limiting U.S. ground action to a few strong
coastal areas (enclaves),” a much larger percentage of respondents was unsure
compared with those who supported a return to bombing (ªgure 4b). Regarding the effects of bombing, ªgures 5a and 5b (which depict 1967 poll results)
show that while slim majorities believed it hampered the North Vietnamese
war effort, a stunning 86 percent (with very few “not sure” responses) believed
that the bombing “backed up our troops in the ªeld.”45
shielding the public from the costs it cared about most
Cost distribution theory suggests that one reason for public hawkishness in
the face of limited success was the war’s relatively low cost for the average
voter. This phenomenon is demonstrated by two sets of polls that distinguish
between war costs that the American people cared about (manpower and casualties) and those they experienced directly (inºation and taxes). If the government is sensitive to voter preferences, responses in these two areas should
diverge as leaders choose strategies that minimize the costs that most concern
voters in favor of those that do not.
Questions in a 1967 poll distinguished between what “troubled” and what
“affected” respondents, showing that only 44 percent of respondents felt that
45. Harris Collection, No. 1623, June 1966; and Harris Collection, No. 1702, January 1967.
International Security 34:3 136
Figure 5. American Public Opinion on U.S. Bombing Effects, January 1967
(a) Agree or disagree? Bombings don’t hurt
North Vietnamese war effort.
(b) Agree or disagree? Bombings back up
our troops in the field.
SOURCE: Survey by Louis Harris and Associates, “Harris 1967 Survey, No. 1702,” in Harris
Collection, Odum Institute, Louis Harris Data Center, University of North Carolina, http://
arc.irss.unc.edu/dvn/dv/odvn/faces/StudyListingPage.xhtml?mode!1&collectionId!4.
their personal lives had been “affected” by the war. Among those affected,
more respondents (32 percent) cited inºation than casualties (25 percent) as the
source. However, responding to the question, “What two or three things about
the war in Vietnam most trouble you personally?” 31 percent said the equivalent of casualties or killing; 12 percent said lack of progress; and only 7 percent
said rising cost. The same questions were asked in March 1968, immediately
after the high U.S. casualty rates resulting from the Tet offensive. More than
half of the respondents thought that the war had affected them personally, and
half of these identiªed inºation and taxes as the principal source. Although
only 9 percent knew someone who had been killed in Vietnam, “concern” over
the drafting of a son or husband had risen to 37 percent. As for the war’s other
“troubling aspects,” 44 percent cited U.S. casualties, and 7 percent cited
ªnancial costs.46
Although there was little difference among income groups regarding approval of U.S. strategy choices in Vietnam, some survey questions allow for
testing the war effort’s distributive elements. Cost distribution theory suggests
that lower-income groups should be less sensitive to the costs of war when ex46. Harris Collection, No. 1734, July 1967; and Harris Collection, No. 1813, March 1968, quoted in
Mark Lorell and Charles Kelley Jr., “Casualties, Public Opinion, and Presidential Policy during the
Vietnam War,” No. R-3060-AF (Santa Monica, Calif.: Project Air Force, RAND, March 1985),
pp. 23–28.
The Myth of Military Myopia 137
pressed in the form of higher taxes than in the form of higher labor costs (i.e.,
conscription). Although few respondents to a January 1967 poll supported an
income tax increase, majorities among those initially opposed became more
supportive if “convinced it would help pay for the war.”47 Moreover, one’s
income appears to have affected this support, as illustrated in ªgure 6.48
Wealthier respondents were more likely to remain opposed to raising taxes.
The relationship to income changes in ªgure 7, when respondents were asked
if they favored shifting the selection of conscription to a lottery system without
deferments. Opposition to a lottery grew with income, suggesting a belief by
many that the deferment system favored the wealthy. This suggests that the
conscription “tax” was considered regressive, making the use of capital that
much more appealing to relatively poor voters.49 Thus, relative income appears to have played a role in public support for the Johnson administration’s
Vietnam policy when costs are made explicit.
government responsiveness to the public
Public opinion polls occasionally directly inºuenced the choice of military
strategy in Vietnam; National Security Adviser McGeorge Bundy, for example,
cited one poll in ªgure 3 in analyzing the differences between the U.S. and
French experiences in Vietnam.50 But often the Johnson administration’s decisions were made in anticipation of public responses. Cost distribution theory
does not require the public to express strong military doctrine preferences.
Rather, it suggests that the government assesses military doctrine in light of
public preferences over outcomes (in the case of Vietnam, a negotiated settlement without full withdrawal) and the costs in blood and treasure.
From the outset, administration ofªcials reiterated the focus on enemy main
47. Harris Collection, No. 1702, January 1967.
48. Ibid. A multinomial logistic regression of respondent’s income on support for the tax increase
among those initially opposed shows that wealthier respondents are more likely to remain so
(with a coefªcient of 0.086 and standard error of 0.05). The wealthier are less likely to answer
“not sure” as well ("0.229, standard error of 0.10). Wealthier respondents are less likely to change
their initial opposition to the income tax, answering “more in favor” or “not sure” (with a coefªcient of "0.107, standard error of 0.05). In the publicly available polling data sets, there are surprisingly few questions linking the war and raising taxes. Generally, polls that addressed income
taxes also asked about a variety of alternate measures, such as price controls, which relatively less
wealthy people tended to favor. See, for example, Gallup Poll, October 21–26, 1966.
49. A multinomial logistic regression of respondent’s income on support for the lottery gives a
coefªcient of "0.087 and standard error of 0.047; that is, the wealthier appear less likely to support
switching to a lottery system. The wealthy were much less likely to answer “not sure” as well
(coefªcient of "0.417, standard error of 0.069).
50. Foreign Relations of the United States (FRUS), 1964–1968, Johnson Administration, Vol. 3, Doc. 33,
http://www.state.gov/www/about_state/history/vol_iii/index.html. For the poll, see Harris
Collection, No. 1331, June 1965.
International Security 34:3 138
Figure 6. U.S. Support for Income Tax Increase “if you were convinced it would help
pay for the war in Vietnam,” by Total Family Income for 1966
SOURCE: Survey by Louis Harris and Associates, “Harris 1967 Survey, No. 1702,” in Harris
Collection, Odum Institute, Louis Harris Data Center, University of North Carolina, http://
arc.irss.unc.edu/dvn/dv/odvn/faces/StudyListingPage.xhtml?mode=1&collectionId=4.
force units and explicitly rejected the use of personnel (especially U.S. forces)
to pursue paciªcation. Recently sworn in and already preparing for the 1964
election, President Johnson convened an ad hoc committee on the conºict,
chaired by diplomat William Sullivan, who had been told in advance that the
policy would be a “slow, very slow, escalation” of bombing pressures against
North Vietnam. Sullivan briefed the Joint Chiefs of Staff (JCS) that a military
“presence” without “heavy forces” would maintain public support indeªnitely.51 One contemporary analysis based on anonymous interviews of State
Department ofªcials assessed that the president was “more inclined to listen”
to advocates of “selective bombing of North Vietnamese targets and clandes51. H.R. McMaster, Dereliction of Duty: Lyndon Johnson, Robert McNamara, the Joint Chiefs of Staff,
and the Lies That Led to Vietnam (New York: HarperCollins, 1997), p. 68.
The Myth of Military Myopia 139
Figure 7. Responses to “Would you have the present draft system or a lottery system
for the draft?” by Total Family Income for 1966
SOURCE: Survey by Louis Harris and Associates, “Harris 1967 Survey, No. 1702,” in Harris
Collection, Odum Institute, Louis Harris Data Center, University of North Carolina, http://
arc.irss.unc.edu/dvn/dv/odvn/faces/StudyListingPage.xhtml?mode=1&collectionId=4.
tine naval raids along the coast” because 1964 was an election year, and
Johnson knew he had to “take some action soon to show that his administration was on top of the situation.”52 This article shows that throughout this
period, the Johnson administration perceived strong constraints from the
public and responded with a conventional military approach to an insurgency, epitomized in a January 1966 phone call between Secretary of Defense
Robert McNamara and President Johnson, in which McNamara reported that
Vietcong defectors “indicate that the pressure that is being applied to them by
air and by constant offensive probing by the [South Vietnam] government and
U.S. forces is beginning to appear in morale.” The defense secretary warned
that “the longer you extend the pause [in strategic bombing of the North], the
more dangerous a [midterm election] campaign issue it becomes.”53
52. Stempel, “Policy/Decision Making in the Department of State,” p. 221.
53. FRUS, 1964–1968, Vol. 4, Doc. 26.
International Security 34:3 140
U.S. Strategy in Vietnam: “Expensive in Dollars, but Cheap in Life”
In the following empirical sections, I discuss why the U.S. government continued to pursue its counterinsurgency strategy despite information that it was
failing. First, I show the overarching philosophy of the war’s prosecution: the
inefªcient substitution of capital for labor. I then show how the Johnson administration chose airpower, barrier construction, and a search-and-destroy/
attrition strategy as its preferred means of ªghting the insurgent component of
the war. Using evidence from internal deliberations and public statements, I
support the claim that civilians made these decisions with domestic politics in
mind.
Toward the end of 1966, Secretary of Defense McNamara told General
Westmoreland that “he would approve whatever related requirements were
developed to ensure that RVN [Republic of Vietnam, also referred to in primary documents as GVN and SVN] manpower and U.S. money substitute for
U.S. blood.”54 Despite obvious evidence of a failing strategy, the Pentagon
Papers (the Defense Department analysis of the war’s conduct) describes further attempts to de-emphasize labor, “certain ‘oblique alternatives,’ those
which were not directly substitutable options” and were “designed to relieve
pressure on U.S. resources, especially manpower.”55 The Johnson administration and its uniformed subordinates highlighted this substitution for the public. Robert Komer, Johnson’s principal counterinsurgency adviser, recalled the
political exchange rate in 1982: “What it costs you in blood is much more politically visible than what it costs you in treasures.”56 JCS Chairman Gen. Earle
Wheeler told members of a Rotary Club, “The United States policy is to expend money and ªrepower, not manpower, in accomplishing the purpose of
the nation.”57 Early in the war, McNamara refused to answer directly a New
York Times writer’s question, “How large a commitment of men is the United
States prepared to make at the end of 1965?” The defense secretary instead replied that “the thing we prize most deeply is not money but men. . . . It’s expensive in dollars, but cheap in life.”58 Responding in a televised congressional
hearing to a question on the economic costs of strategic bombing, McNamara
stated that ªnancial comparisons do not “have great value in affecting the de54. Gibbons, The U.S. Government and the Vietnam War, Vol. 4, pp. 106–107.
55. Mike Gravel, ed., The Pentagon Papers: The Defense Department History of United States
Decisionmaking on Vietnam, Vol. 4 (Boston: Beacon, 1971), p. 385.
56. Lorell and Kelley, “Casualties, Public Opinion, and Presidential Policy during the Vietnam
War,” p. 80.
57. Quoted in Krepinevich, The Army and Vietnam, p. 198.
58. Henry F. Graff, The Tuesday Cabinet: Deliberation and Decision on Peace and War under Lyndon B.
Johnson (Englewood Cliffs, N.J.: Prentice-Hall, 1970), pp. 81–82.
The Myth of Military Myopia 141
cisions as to whether to bomb or not bomb speciªc facilities.” Rather, “one of
the standards I use in recommending targets for attack on the North [is comparing] the value of facilities destroyed in the North with the number of U.S.
lives lost in the process of destroying them.”59
The principal means of substituting capital for labor was the employment of
air- and artillery-delivered ordnance to a degree that counterinsurgency expert
Robert Thompson observed, “All ground operations were designed to achieve
a ªx on an enemy unit so that every modern weapon could be brought to
bear.”60 Ordnance was expended routinely, even when it was likely to have
only a marginal effect; 70 percent of artillery ªre was employed in situations of
light or inactive combat intensity.61 Upon hearing McNamara’s “reluctant” endorsement of Westmoreland’s initial request to deploy ground forces in 1965,
Johnson desperately sought alternatives, asking McNamara if “there’s any
way, Bob, that through your small planes or helicopters . . . you could spot
these people and then radio back and let the planes come in and bomb the hell
out of them?” The president even suggested getting “every damn admiral that
we’ve got that want some practice,” to bring the U.S. Navy to bear on the
conºict.62
airpower as counterinsurgency strategy
The Johnson administration designed both the strategic bombing of North
Vietnam and the more tactically oriented air operations in the South with the
insurgency in mind.63 A pivotal memo on “sustained reprisal,” written by
National Security Adviser Bundy in February 1965, argued that strategic
bombing was a “new norm in counter-insurgency,” because “to stop it [the
bombing] the Communists would have to stop enough of their activity in the
South to permit the probable success of a determined paciªcation effort.”64
Later that year, Bundy noted the importance of population security and “the
civil side of the war,” but then described how it should be “fought in the shelter of sea and air power.”65 Walt Rostow, Bundy’s successor as national secu59. Preparedness Investigating Subcommittee, Armed Services Committee, U.S. Senate, Air War
against North Vietnam, 90th Cong., 1st sess. (Washington, D.C.: U.S. Government Printing Ofªce,
1967), p. 283.
60. Robert G.K. Thompson, No Exit from Vietnam (New York: D. McKay, 1969), p. 135.
61. Thomas C. Thayer, War without Fronts: The American Experience in Vietnam (Boulder, Colo.:
Westview, 1985), p. 57.
62. Quoted in Michael Beschloss, Reaching for Glory: Lyndon Johnson’s Secret White House Tapes,
1964–1965 (New York: Simon and Schuster, 2002), pp. 194–195.
63. Kocher, Pepinsky, and Kalyvas, “Aerial Bombardment, Indiscriminate Violence, and Territorial
Control.”
64. FRUS, 1964–1968, Vol. 2, Doc. 84.
65. Graff, The Tuesday Cabinet, p. 94.
International Security 34:3 142
rity adviser in 1966, agreed, describing airpower as “the equivalent of guerilla
warfare.”66
The administration understood that bombing was unlikely to be effective,
even as it ordered its use. In a late-1964 memo to Maxwell Taylor, the U.S. ambassador to Vietnam and a key presidential military adviser, Johnson observed, “I have never felt that this war will be won from the air,” making the
point more colorfully in a March 1965 phone conversation with Senator
Richard Russell, “Airplanes ain’t worth a damn Dick!”67
The United States nonetheless embarked on a massive air campaign and
continued it long after sufªcient evidence existed that it had little, if any, effect
on the insurgency. The near one-to-one ratio of combat sorties in the theater to
personnel in Vietnam is shown in ªgure 8. In 1966 the number of sorties overtook the number of personnel, rose more steeply to the peak deployment of
1968, and then declined less sharply. Before the ratio increased by an order of
magnitude in 1972, the largest sorties-to-personnel ratio occurred during 1969,
when the U.S. military had supposedly shifted to a less ªrepower-intensive
paciªcation strategy. Administration ofªcials set daily requirements for ordnance expenditure and air strikes, causing Westmoreland to complain in an
eyes-only cable to Paciªc Forces commander Adm. Ulysses Sharp that
McNamara “manifested uncommon interest” about reports that these quotas
were not met by available in-country resources. McNamara “voiced concern
that ‘policy’ concerning use of in-country and carrier based air in support of
operations in SVN [was] not being adhered to,” stressing the “relative ease” of
obtaining “additional carriers if needed.”68
President Johnson was much less worried about revealing to the public the
extent of the bombing campaign than he was the ground escalation.69 National
Security Action Memorandum (NSAM) 328, authorizing expansion of the
ground and air campaign in April 1965, infamously noted the president’s desire to “minimize any appearance of sudden changes in policy,” but this admonition applied only to the additional personnel deployments and the “more
active use” of Marines in Vietnam. Other actions, in particular “the present
66. Walt Rostow to Averell Harriman, January 28, 1966, Box 13, Papers of Walt Rostow, Lyndon
Baines Johnson (LBJ) Library, quoted in David Milne, “‘Our Equivalent of Guerilla Warfare’: Walt
Rostow and the Bombing of North Vietnam, 1961–1968,” Journal of Military History, Vol. 71, No. 1
(January 2007), pp. 169–203.
67. FRUS, 1964–1968, Vol. 1, Doc. 477; and Beschloss, Reaching for Glory, p. 212 (emphasis in
original).
68. Westmoreland to Sharp, July 17, 1965, Eyes Only Message File (EOMF), Box 34, LBJ Library.
McNamara was constantly pushing Wheeler and Westmoreland to improve the close air support
provided to army ground forces. Wheeler to Sharp, December 22, 1965, EOMF, Box 30.
69. Herbert Y. Schandler, The Unmaking of a President: Lyndon Johnson and Vietnam (Princeton, N.J.:
Princeton University Press, 1977), p. 61.
The Myth of Military Myopia 143
slowly ascending tempo of ROLLING THUNDER operations [i.e., strategic
bombing of North Vietnam],” were not subject to this restriction.70 When
McNamara advised extending by three days the May 1965 pause in bombing
to “satisfy the New York Times editorial board,” Johnson responded, “If we hold
off bombing any longer, people are going to say ‘What in the world is happening?’ My judgment is the public never wanted us to stop the bombing.”71
In an October 1965 memo, Assistant Secretary of State for Far Eastern Affairs
William Bundy linked strategic bombing, U.S. casualties, and public opinion:
“We are faced with the pressures from various quarters . . . to hit the
North substantially harder. The degree to which this will rise during the next
3–6 months will depend heavily on actual casualty experience.” Bundy tied
support for increased bombing to the “Phase II” ground force deployment anticipated for 1966 in light of “US domestic reaction.” Bundy cautioned, “There
would be a lot of rumbling below decks and among the harder-action school of
critics,” and “pressures would be enormous thereafter [a bombing pause]
to ‘really clobber’” the North.72 In the November 1967 deliberation over
McNamara’s proposal to level off ROLLING THUNDER sorties, Rostow observed that “acknowledging my limitations as a judge of domestic politics, I
am extremely skeptical of any change in strategy that would take you away
from your present middle position.” Rostow argued, “If we shift unilaterally
to de-escalation, the Republicans will move in and crystallize a majority toward a stronger policy.”73 Taylor attached a note to Rostow’s analysis concurring that the curtailment of U.S. bombing would mobilize “the large majority
of our citizens who believe in the bombing but who thus far have been
silent.”74
The popularity of airpower, its linkage to troop levels and casualties in the
minds of the American public and policymakers, and the limits of military
inºuence on the administration are well illustrated by a rare instance of successful military subversion of presidential policy. Given the opportunity in the
summer of 1967 to testify publicly before John Stennis’s hawkish Military
Preparedness Subcommittee, the JCS pushed for an expansion of U.S. strategic
bombing even as the defense secretary, convinced by then of ROLLING
THUNDER’s futility, resisted. McNamara gave a masterful brief on bombing’s
70. FRUS, 1964–1968, Vol. 2, Doc. 242.
71. May 16, 1965, NSC Meeting Notes, Box 1, LBJ Library, quoted in McMaster, Dereliction of Duty,
pp. 284–285.
72. FRUS, 1964–1968, Vol. 3, Doc. 181. For similar sentiments from Rostow and McNamara, see
FRUS, 1964–1968, Vol. 4, Doc. 232, which refers to the November elections.
73. FRUS, 1964–1968, Vol. 5, Doc. 381.
74. Quoted in Gibbons, The U.S. Government and the Vietnam War, p. 887.
International Security 34:3 144
limitations, but he was contradicted by both the generals and the senators,
who suggested that “we probably would have suffered fewer casualties in the
south if the air campaign against the north had not been burdened with restrictions and prohibited targets.”75
As but one example, Senator Stuart Symington highlighted the substitution
logic: “If instead of diminishing the bombing effort you increased it, you could
do as good a job or a better job, with less troops in South Vietnam could you
not?”76 But Symington also revealed an awareness of the bombing’s ineffectiveness—“Why is it that we are putting out this gigantic effort, but getting so
little, so terribly little results? It is what everybody wants to talk about when I
go back to Missouri”—and an unwillingness to consider another strategy—
“The people are now beginning to realize that we have shackled our seapower
and shackled our airpower.”77 Despite his dismay over JCS maneuvering,
President Johnson began to abandon his civilian advisers’ recommendations in
favor of the military’s more politically palatable ones.78
counterinsurgency by barrier
Even when the air war’s shortcomings became clear, the Johnson administration chose neither to increase the number of ground forces nor to reconsider
U.S. involvement in Vietnam. Instead it attempted to build one of the Pentagon
Papers’ “oblique strategies”: a collection of electronic surveillance equipment,
mines, and physical barriers to prevent inªltration into the South. Variously
named “Practice Nine,” “Muscle Shoals,” and “Igloo White,” its colloquial
name became the “McNamara Line,” reºecting the defense secretary’s enthusiastic support. Designed partly to head off an army request for four more divisions to block incursions from the North, its development was championed by
McNamara and other politically oriented ofªcials in the face of objections from
military experts. The Joint Chiefs generally agreed with Admiral Sharp’s assessment: “an inefªcient use of resources with small likelihood of achieving
U.S. objectives in Vietnam.”79
Despite these objections, McNamara approved the plan in September 1966,
including it in a presidential memo as one of the ªve principal means of reversing the war’s course, estimating that construction would cost $1 billion
with an annual operating cost of $800 million.80 In early 1967, NSAM 358 as75.
76.
77.
78.
79.
80.
Preparedness Investigating Subcommittee, Air War against North Vietnam, p. 208.
Ibid., p. 34.
Ibid., p. 420.
Schandler, The Unmaking of a President, p. 61.
Gravel, The Pentagon Papers, Vol. 4, p. 114.
FRUS, 1964–1968, Vol. 4, Docs. 233 and 268.
The Myth of Military Myopia 145
signed the “highest national priority category” to the program.81 Faced with
JCS foot-dragging, McNamara shouted, “Get on with it for God’s sakes, it’s
only money!”82
Not only was the barrier given near-limitless resources, but its development
was designed with public opinion in mind. In a May 19, 1967, memo on
ªnding a way to reduce bombing in a manner “acceptable to our own people,”
National Security Adviser Rostow mooted, “Surfacing the concept of the barrier may be critical to that turnaround [in public opinion].”83 Polls backed up
this assessment; when respondents were asked to evaluate ways “to step
up our military effort in Vietnam,” the most popular option (60 percent in support vs. 18 percent opposed) was “building a military barrier across all entrance routes into South Vietnam.”84 A Washington Post column trumpeted the
(leaked) project as “a revolutionary new approach” that could “conceivably
transform the Vietnamese war.”85 Originally proposed by the Pentagon’s
JASON advisory group as a substitute for the ineffective air campaign, the barrier simply supplemented it.86
main force focus as counterinsurgency strategy
One of the Vietnam War’s most puzzling aspects, and an important component of the military myopia case, is the leeway the Johnson administration apparently gave to the military in pursuing an ultimately unsuccessful ground
strategy.87 Given civilian micromanagement elsewhere, Occam’s razor suggests a simple explanation: the commander in the ªeld was doing precisely
what the president wanted him to do.88 Johnson understood what type of war
would be fought, and throughout the conºict he and his advisers reinforced
the pursuit of this strategy.
Both civilian and military leaders regarded the setting of ceilings on troop
numbers as sufªcient to preclude a COIN approach.89 A senior deputy’s eyesonly message to General Westmoreland observed that it seemed, “the smaller
81. Walt Rostow, “NSAM 358: Assignment of Highest National Priority to the Mk 84, Mod 1 2000
lb. Bomb and to Project PRACTICE NINE,” LBJ Library, http://www.lbjlib.utexas.edu/johnson/
archives.hom/NSAMs/nsam358.asp.
82. Quoted in David Halberstam, The Best and the Brightest (New York: Modern Library, 2001),
p. 630.
83. FRUS, 1964–1968, Vol. 5, Doc. 162. See also Gravel, The Pentagon Papers, Vol. 4, p. 385.
84. Harris Collection, No. 1735, May 1967.
85. Rowland Evans and Robert Novak, “Inside Report . . . The Vietnam Wall,” Washington Post,
August 1, 1966.
86. Gravel, The Pentagon Papers, Vol. 4, pp. 115–123.
87. Krepinevich, The Army and Vietnam, p. 165.
88. I am grateful to John Nagl for making this point.
89. McMaster, Dereliction of Duty, p. 272.
International Security 34:3 146
the number of maneuver battalions the more B-52’s we need.”90 Yet the U.S.
Army adapted to the constraints in ways inconsistent with the military myopia story, deploying dismounted infantry units in 1965 rather than armored
brigades and divisions, so as to ªeld as many soldiers as possible given personnel caps.91 Moreover, the generals understood that heavy forces were inappropriate for the terrain and the type of war being fought; Westmoreland
insisted to his Pentagon-based colleagues that “Vietnam is no place for either
tank or mechanized infantry units.” Army Chief of Staff Gen. Harold Johnson
agreed: “The presence of tank formations tends to create a psychological atmosphere of conventional combat.”92 The army’s shift to armor occurred gradually over 1966 only as the need to substitute for labor became more apparent.93
administration understanding and rejection of coin. Few summations of effective COIN strategy improve on McNamara’s memo to Johnson in
March 1964, which expresses “the basic theory now fully accepted both on the
Vietnamese and U.S. sides . . . concentrating on the more secure areas and
working out from these through military operations to provide security, followed by necessary civil and economic actions to make the presence of the
government felt and to provide economic improvements.”94 The administration understood the limitations of ªghting an insurgency conventionally;
McNamara briefed the president in July 1965, arguing that “success against the
larger, more conventional, VC/PAVN [People’s Army of North Vietnam]
forces could merely drive the VC back into the trees and back to their 1960–64
pattern—a pattern against which U.S. troops and aircraft would be of limited value.”95 “The large-unit operations war,” McNamara briefed Johnson in
October 1966, “is largely irrelevant to paciªcation as long as we do not lose it.”
Yet the president summarily rejected the COIN option on multiple occasions.
The ground campaign was conªned to using ordnance for main force attrition;
the president constantly exhorted his chiefs to “kill more VC.”96 Despite information suggesting that this approach was not working well, Johnson stuck
with it.
More accurately, the administration’s paciªcation strategy was the main
90. Depuy to Westmoreland, January 31, 1966, EOMF, Box 35.
91. Westmoreland to H. Johnson, July 5, 1965, and July 7, 1965; and Westmoreland to Waters, August 11, 1965, EOMF, Box 34.
92. H. Johnson to Westmoreland, July 3, 1965, EOMF, Box 34.
93. Donn A. Starry, Armored Combat in Vietnam (New York: Arno, 1980), p. 56.
94. FRUS, 1964–1968, Vol. 1, Doc. 84.
95. FRUS, 1964–1968, Vol. 3, Doc. 67.
96. Jack Valenti (one of Johnson’s most loyal aides) observed in an eyes-only memo to the president that “you are rightly judging the trends of the war from . . . numbers of VC killed” and that
“the kill rate is vital to you judging the amount of punishment being meted to the enemy.” Valenti
to Johnson, March 24, 1966, Box 9, White House Central Files, Conªdential File, LBJ Library.
The Myth of Military Myopia 147
force war. During a brieªng on ground force employment on July 21, 1965,
Johnson expressed concern about putting “U.S. forces in those red areas.”
McNamara replied, “You’re right. We’re placing our people with their backs
to the sea—for protection. Our mission would be to seek out the VC in largescale units.”97 In a September 1965 memo to Johnson, National Security
Adviser Bundy summed up the challenge facing the president: deciding “how
we use our substantial ground and air strength effectively against small-scale
harassment-type action, whether we should engage in paciªcation as opposed
to patrolling actively, and whether, indeed, we should taper off our ground
force build-up.” Bundy reported that “we asked [Ambassador Henry Cabot]
Lodge to develop a speciªc plan for our joint consideration which would involve the concentration of GVN forces on paciªcation and the reliance on U.S.
forces to handle large-scale VC actions.”98
In November 1965 deliberations over how best to pursue paciªcation,
Ambassador Lodge forcefully argued that the “crux of the problem” in the
U.S. paciªcation effort “is security. To meet this need we must make more U.S.
troops available to help out in paciªcation operations as we move to concentrate ARVN [Army of the Republic of Vietnam] effort in this work.”99 The
reaction could not have been stronger or clearer: a joint telegram from Rusk,
McNamara, and Komer stated that beyond Westmoreland’s planned “use of
limited number U.S. forces in buddy system principle to guide and motivate
ARVN . . . there should be no thought of U.S. taking on substantial share of
paciªcation.” Rusk later emphasized to Lodge that the strategy was discussed
“at highest levels [i.e., the president], who wished to emphasize that this represents ªnal and considered decision.”100 At this time, McNamara not only recommended troop increases without revising the ground strategy—“The
principal task of U.S. military forces in SVN must be to eliminate the offensive
capability of the regular units”—but gave Johnson a choice between two versions of search and destroy. The ªrst would be “to increase friendly forces as
rapidly as possible, and without limit, and employ them primarily in largescale ‘seek out and destroy’ operations to destroy the main force VC/NVA
97. FRUS, 1964–1968, Vol. 3, Doc. 71.
98. FRUS, 1964–1968, Vol. 3, Doc. 151. Taylor, writing to the president at the end of 1965, not only
recommended that “the role of our U.S. ground forces in this campaign for increased population
security should be primarily the destruction of mainline Viet Cong/North Vietnamese units,” but
criticized the “inclination to turn over all or most of the heavy ªghting to U.S. forces and allow the
bulk of the Vietnamese forces to retire behind a screen of U.S. provided protection to perform
clearing jobs and local defense. . . . At least half of the South Vietnamese regular units should be
used in mobile combat roles.” FRUS, 1964–1968, Vol. 3, Doc. 250.
99. The telegram is found in Gravel, The Pentagon Papers, Vol. 2, pp. 602–605, and footnoted in
FRUS, 1964–1968, Vol. 4, Doc. 290.
100. FRUS, 1964–1968, Vol. 4, Doc. 304
International Security 34:3 148
[North Vietnamese Army] units.” The second was “a similarly aggressive
strategy of ‘seek out and destroy’ but to build friendly forces only to that level
required to neutralize the large enemy units.” McNamara advocated a shift to
the second option in part because “an endless escalation of U.S. deployments
is not likely to be acceptable in the U.S.”101
When Westmoreland requested additional personnel in April 1967, Komer
noted that the enemy main force strength had leveled off and that half of the
U.S. maneuver battalions were already supporting paciªcation “by dealing
with the middle war, the VC main force provincial battalions.” Johnson’s
paciªcation expert warned that a “major U.S. force commitment to paciªcation
also basically challenges the nature of our presence in Vietnam and might force
U.S. to stay indeªnitely in strength.” Observing the political reality that “another major U.S. force increase raises so many other issues,” Komer recommended more Vietnamese involvement, coupled with “a minor force increase
. . . accelerated emphasis on a barrier, and some increased bombing.”102 Even
administration ofªcials favoring escalation evinced modest expectations for
the reinforcements’ effectiveness, concern for the domestic implications, and
no change in strategy. In a May 1967 memo explicitly addressing public responses to escalation strategies, Rostow established the goal of freeing up “additional allied forces to permit Westy to get on with our limited but real role in
paciªcation, notably with the defense of I Corps in the North and the hounding of provincial main force units.”103 Undersecretary of State Nicholas
Katzenbach believed the time had come to “change” the war strategy and “use
the great bulk of U.S. forces for search and destroy.” A “small number” of
troops could be used for paciªcation but “targeted primarily on enemy provincial main force units.”104 During a July 1967 meeting, Johnson agreed with
McNamara’s recommendation that “U.S. units will continue to destroy the enemy’s main force units,” while ordering his subordinates to “shave [any additional troops requests] the best we can.”105
Johnson administration ofªcials clearly based their rejection of COIN on an
anticipation of massive troop requirements and casualties. Responding to junior army ofªcers recommending increased attention to paciªcation, General
Johnson (Army chief and paciªcation proponent) cautioned, “We are not going
to be able to respond to the public outcry in the United States about [the]
101. FRUS, 1964–1968, Vol. 4, Doc. 312.
102. FRUS, 1964–1968, Vol. 5, Doc. 147.
103. FRUS, 1964–1968, Vol. 5, Doc. 162.
104. Gravel, The Pentagon Papers, Vol. 4, p. 508.
105. “Notes from Meeting of the President with Secretary McNamara to Review the Secretary’s
Findings during Vietnam Trip,” July 12, 1967, Tom Johnson Papers, Box 1, LBJ Library.
The Myth of Military Myopia 149
casualties that might result.”106 In a July 1965 memo to President Johnson,
McNamara noted the differences in manpower requirements between main
force and counterinsurgency operations: “The number of U.S. troops is too
small to make a signiªcant difference in the traditional 10-to-1 governmentguerrilla formula.”107 According to Katzenbach, “Paciªcation is not the ultimate answer—we have neither the time nor the manpower.”108
civilians overriding military coin recommendations. The military
myopia interpretation of Vietnam is based on the U.S. armed forces pushing a
ºawed strategy up the chain of command or the existence of an environment
of neglect allowing the U.S. military to pursue its problematic aims. The record
does not support this interpretation.
Not only did U.S. military leaders generally give in to Johnson’s demands,
but they also calibrated their recommendations to make them as politically
palatable as possible.109 Asked if domestic opposition was a factor in JCS decisions, General Wheeler replied, “Not directly . . . [but] the Chiefs are well
aware of the problems engendered for the President by the minority dissent to
his course of action,” citing the need not to “put a club in the hands of dissenters.”110 When the JCS were deemed unlikely to support the president’s strategy, they were excluded from deliberations. Many of the most important
military policies—the escalation decisions of June and July 1965, the establishment of the principal war aim of “killing more VC,” the emphasis on B-52
bombing of Vietcong sanctuaries—were made with little strategic input from
the JCS, to the point of lying to Wheeler about the purposes of meetings to
which he was not invited.111 Only when McNamara and other civilian ofªcials
turned against the prevailing (and politically popular) airpower strategy did
the Joint Chiefs ªnd a more receptive presidential audience.
Indeed, General Westmoreland was one of the few in 1964 to recommend
continuing the advisory effort—“Option A” of the famous three-option framework that led to the ROLLING THUNDER bombings.112 In a January 6, 1965,
cable to Johnson via Ambassador Taylor, Westmoreland asserted that “if
[the U.S. advisory] effort has not succeeded there is less reason to think that
106. Krepinevich, The Army and Vietnam, p. 172.
107. FRUS, 1964–1968, Vol. 3, Doc. 67.
108. Gravel, The Pentagon Papers, Vol. 4, p. 508.
109. McMaster, Dereliction of Duty, especially chap. 15.
110. Graff, The Tuesday Cabinet, pp. 125–126.
111. McMaster, Dereliction of Duty, p. 301.
112. For a description of this strategy’s development, see Gravel, The Pentagon Papers, Vol. 3,
pp. 205–251. On Westmoreland’s understanding of the importance and proper conduct of
paciªcation, see Andrew J. Birtle, “PROVN, Westmoreland, and the Historians: A Reappraisal,”
Journal of Military History, Vol. 72, No. 4 (October 2008), pp. 1213–1247.
International Security 34:3 150
U.S. combat forces would have the desired effect. . . . Intervention with ground
combat forces would at best buy time and would lead to ever increasing commitments.” Westmoreland argued that instances from 1963 to 1964 where U.S.
ground forces would have been helpful were “few and far between. . . . In balance, they do not seem to justify the presence of U.S. units.”113 In an eyes-only
message to Wheeler, Westmoreland was even more emphatic, recommending
the “present policy” until “some positive momentum in paciªcation” was
made. “Expanded and concerted U.S. attacks” on North Vietnam were inadvisable until justiªed by “a ªrmer RVN base and prospects for victory.” He
continued that South Vietnam’s government “may become unhealthily preoccupied with external operations to the detriment of paciªcation.”114
In September 1965 National Security Adviser Bundy reported to Johnson
that “Lodge and Westmoreland feel VC ‘lie-low’ tactics will become increasingly a police-social action problem” and summarized Westmoreland’s strategy before recommending its rejection: “Destroy VC units where they can be
found and pacify selected high priority areas, restore progressively the entire
country to GVN control, support ‘rural reconstruction’ with comprehensive attention to the paciªcation process.”115 Admiral Sharp, in an eyes-only message
to Wheeler, complained that the Department of State “is somehow hopeful
paciªcation may be achieved by the Vietnamese themselves while being aided
by little if any U.S. participation.” Sharp continued, “We will do far better in
paciªcation if we too press forward setting the example in performance and results. . . . The GVN cannot do the paciªcation alone, this would prolong the
struggle beyond foreseeable limits. If the Viet Cong go underground and revert to small-scale actions, we should employ U.S. forces in coordination with
the ARVN and proceed with securing and pacifying areas as fast as we can.”116
Contradicting both McNamara’s recommendations and the claims of the
military myopia argument, Westmoreland proposed a new concept of operations in August 1966 that devoted “a signiªcant number of the U.S./Free
World Maneuver Battalions” to paciªcation missions, which “encompass
base security and at the same time support revolutionary development by
spreading security radially from the bases to protect more of the population.
Saturation patrolling, civic action, and close association with ARVN, regional
and popular forces to bolster their combat effectiveness are among the tasks of
the ground force elements.” In an accompanying memo to President Johnson,
113.
114.
115.
116.
FRUS, 1964–1968, Vol. 2, Doc. 13.
Westmoreland to Wheeler, “Future U.S. Actions in RVN,” November 26, 1964, EOMF, Box 30.
FRUS, 1964–1968, Vol. 3, Doc. 151.
Sharp to Wheeler, September 22, 1965, EOMF, Box 34.
The Myth of Military Myopia 151
endorsed by Rostow, Ambassador Taylor acknowledged that Westmoreland’s
strategy could result in “speeding up the termination of hostilities,” but he
cautioned that “there will be a cost to pay for this progress in a rise in the U.S.
casualty rate.” After noting the likely negative reaction domestically, he concluded that if paciªcation became the strategy, “General Westmoreland will be
justiªed in asking for almost any ªgure in terms of future reinforcements.” A
handwritten note on a follow-up memo cites Johnson’s instructions to “get
something to Westy so that he will not assume that we have approved.”117
Civilian objections to military recommendations continued into the next
year. In a May 1967 memo arguing that “the ‘philosophy’ of the war should be
fought out now,” John McNaughton, assistant secretary of defense for international security affairs and an important adviser on U.S. strategy in Vietnam,
counseled the rejection of Westmoreland’s March 1967 request for 200,000
more soldiers, because Westmoreland intended to use the reinforcements “to
relieve the Marines to work with ARVN on paciªcation” and for similar missions in the Mekong Delta and Quang Ngai Province. Referring to paciªcation
as a “less essential mission,” McNaughton suggested avoiding escalation
by “making more efªcient use of presently approved U.S. manpower (e.g.,
by removing them from the Delta, by stopping their being used for paciªcation work in I Corps, by transferring some combat and logistics jobs to
Vietnamese or additional third-country personnel).”118 McNamara agreed
with McNaughton’s recommendation in a memo to Johnson that reveals the
sensitivity to troop deployment and a continued main force focus: “We will
soon have in Vietnam 200,000 more U.S. troops than there are in enemy main
force units. We should therefore, without added deployments, be able to maintain the military initiative, especially if U.S. troops in less-essential missions
(such as in the Delta and in paciªcation duty) are considered strategic reserves.” McNamara justiªed the rejection in part by noting that Westmoreland
intended to use the bulk of the ªrst 100,000 troops for paciªcation.119 The focus
on conventional combat was not limited to the Pentagon; Westmoreland complained to Wheeler and Sharp in August 1967 that “Secretary Rusk is thinking
in terms of the more conventional type warfare where our forces could launch
such an all-out offensive from a reasonably secure area of departure, leaving
behind a paciªed rear area, and against identiªed enemy formations disposed
along a recognizable front. Such is not the case in SVN.”120
117. FRUS, 1964–1968, Vol. 4, Docs. 220–223. Taylor’s analysis in document 221 explicitly describes
and rejects every aspect of COIN recommended by Westmoreland.
118. FRUS, 1964–1968, Vol. 5, Doc. 161.
119. FRUS, 1964–1968, Vol. 5, Doc. 177.
120. Westmoreland to Wheeler and Sharp, August 26, 1967, EOMF, Box 37.
International Security 34:3 152
civilian rejection of the marines’ combined action program. Military
myopia claims largely rest on the counterfactual that an extension of the
Marines’ innovative Combined Action Program (CAP), the conºict’s best example of American COIN, could have employed the same number of soldiers
stationed in Vietnam while minimizing casualties and enhancing population
security.121 Obviously, this argument cannot be tested, but it is also irrelevant if
civilian as well as military leaders did not hold this position at the time.122
Andrew Krepinevich’s claim that a force of 167,000 U.S. soldiers was sufªcient
to blanket South Vietnam with CAP teams is based on the reports of the
Pentagon’s Systems Analysis Ofªce (the “SEA reports”).123 These reports did
lambaste the prevailing attrition strategy and acknowledged CAP’s excellence,
but they were also skeptical of CAP’s wider viability.
According to the SEA reports, broadening CAP required 279,000 additional Popular Forces (PF) militia members, and thus “the reluctance of the
[South Vietnam government] to assign PF personnel to CAPS is a serious problem in considering any expansion.” Between July 1967 and November 1968,
the PF-to-Marine ratio had declined from 1.7 to 1.4.124 As of mid-1967, SEA assessed that a CAP Marine had a 75 to 80 percent chance of being wounded and
a 16 to 18 percent chance of being killed.125 Finally, a November 1968 SEA report observed, “In over three years of operations no evidence exists that U.S.
Marines have been able to withdraw from a CAP solely because their Vietnamese counterparts were able to take over.”126
Although Krepinevich dismisses as “lip service” Westmoreland’s objection
to CAP—“I simply had not enough numbers to put a squad of Americans
in every village and hamlet”—this assessment was shared by an administration determined to hold down deployments and casualties. Indeed, Ambassador Taylor found Westmoreland too receptive to “the ‘oil spot’ concept as the
121. Krepinevich, The Army and Vietnam, p. 176; Nagl, Learning to Eat Soup with a Knife; and
Thompson, No Exit from Vietnam, p. 198. A CAP team consisted of a thirteen-man Marine riºe
squad assigned to a local thirty-ªve-man Vietnamese militia platoon living among the people to
provide both security and civil assistance.
122. As I emphasize throughout this article, determining the circumstances under which a solid
COIN approach results in fewer casualties remains an important policy question.
123. A more modest estimate in early 1969 suggested that given the number of troops deployed,
the CAP concept could be extended to 2,500 (of 12,000) hamlets. Hugh Hanning, ed., Lessons from
the Vietnam War (London: Royal United Service Institution, 1969), p. 18.
124. Thomas C. Thayer, ed., Paciªcation and Civil Affairs: A Systems Analysis View of the Vietnam War,
1965–1972, Vol. 10 (Washington, D.C.: Ofªce of the Assistant Secretary for Defense [Systems Analysis], 1975), pp. 26–27; and Jack Schulimson, U.S. Marines in Vietnam: 1968, The Deªning Year
(Washington, D.C.: History and Museums Division, U.S. Marine Corps, 1997), p. 628.
125. Michael E. Peterson, The Combined Action Platoons: The U.S. Marines’ Other War in Vietnam
(New York: Praeger, 1989), pp. 87–88.
126. Thayer, Paciªcation and Civil Affairs, p. 13.
The Myth of Military Myopia 153
Marines have been doing in the I Corps area (and other U.S. forces elsewhere
to a lesser degree).”127 Seeking ways to trim personnel in 1967, McNaughton
noted, “Other ground-force requirements could be eliminated if the U.S.
Marines ceased grass-roots paciªcation activities.”128 Interviewed in 1976,
Komer assessed that CAP demanded “an enormous requirement for American
infantry which we did not have.”129 Written in mid-to-late 1968, the Pentagon
Papers acknowledges CAP’s unquestioned success relative to any other approach but warns that the Marine strategy “requires vast numbers of troops,”
and should only be “undertaken with full awareness by the highest levels of
the [U.S. government] of its potential costs in manpower and time.”130 It is unrealistic to think that the Johnson administration would have supported such a
program given the political limitations on personnel and casualties and the
easy availability of matériel. A competent COIN strategy along the lines of
CAP might have been more effective and reduced casualties, but the civilian
leadership was unwilling to take that chance. William Bundy wrote to
Katzenbach that paciªcation of the Mekong Delta region should be avoided, as
“apart from the military merits, any force increase that reaches the ‘Plimsoll
Line’—calling up the Reserves—. . . might also lead to pressures to go beyond
what is wise in the North, speciªcally mining Haiphong.”131
did nixon and abrams ªght a better war?
Some accounts claim that once Gen. Creighton Abrams replaced
Westmoreland as MACV commander in mid-1968, the war began to be
fought successfully.132 Yet any changes in tactics on the ground, such as
Vietnamization, were driven by the decisions of Presidents Johnson and Nixon
to freeze and then lower the level of U.S. troop deployments. Indeed, U.S. military doctrine remained consistent, and control of the military remained ªrmly
in the hands of administration ofªcials.
Although reductions in U.S. military personnel transferred much responsibility for the ªghting to the Vietnamese, the U.S. contribution to counterinsurgency retained its ªrepower-intensive emphasis. Figure 8 shows that the gap
between sorties and personnel reached its maximum in the ªrst two years of
Abrams’s command, after the U.S. “shift” to a paciªcation strategy. Consump-
127. FRUS, 1964–1968, Vol. 4, Doc. 221.
128. FRUS, 1964–1968, Vol. 5, Doc. 161.
129. Quoted in Schulimson, U.S. Marines in Vietnam, p. 620.
130. Gravel, The Pentagon Papers, Vol. 2, p. 535.
131. FRUS, 1964–1968, Vol. 5, Doc. 154.
132. Sorley, A Better War. See also Nagl, Learning to Eat Soup with a Knife, pp. 168–174; and Gray, Irregular Enemies and the Essence of Strategy, pp. 17–18.
International Security 34:3 154
Figure 8. Numbers of U.S. Personnel/Aircraft Sorties in Vietnam/Southeast Asia,
1965–72
SOURCE: Thomas C. Thayer, War without Fronts: The American Experience in Vietnam (Boul-
der, Colo.: Westview, 1985), p. 32.
tion of artillery rounds remained constant from June 1967 to June 1970, even as
200,000 troops were drawn down.133 General Abrams kept his air cavalry units
in theater until 1972 to provide “a maximum of ªrepower and mobility with
a minimum of U.S. troops.” By the end of 1971, armored units represented
more than half of the U.S. maneuver battalions still in Vietnam.134 Removing
infantry ªrst lowered casualty rates but at the cost of diminished COIN
effectiveness.135 Komer did not believe that the new commander altered
Westmoreland’s strategy: “There was no change in strategy whatsoever. In fact
[Abrams] said he didn’t intend to make any changes unless he saw that some
were necessary.” Instead, the strategy “didn’t really change until we began
withdrawing.”136
133. Thayer, War without Fronts, p. 57.
134. Starry, Armored Combat in Vietnam, pp. 164–165.
135. Thayer, War without Fronts, p. 122.
136. W. Scott Thompson and Donaldson D. Frizzell, eds., The Lessons of Vietnam (New York: Crane,
Russak, 1977), pp. 79–93. In this conference report, several veteran Vietnam counterinsurgency
The Myth of Military Myopia 155
This is not to blame Abrams; he had little control over numbers set by the
president and Congress, whose priorities are clear from budget and deployment ªgures. Of the $21.5 billion spent in ªscal year 1969, only 5 percent went
toward paciªcation and civil operations; this ratio remained the same through
1971. Thomas Thayer, in charge of the SEA reports for this period, describes
the U.S. war effort of 1969–71 as “ªrst and foremost an air war although
Vietnam was billed as a land war in Asia, and second, a ground attrition campaign against communist regular units. Paciªcation was a very poor third.”137
Conclusion
This article shows that, contrary to the consensus regarding U.S. military intransigence in the face of unconventional warfare, civilian ofªcials in Lyndon
Johnson’s administration—and ultimately the American public—played an essential role in the selection of a capital-intensive strategy to ªght insurgents in
the Vietnam War. President Johnson was convinced that the American public
would punish any administration that “lost” South Vietnam to communism,
but he was equally certain that public preferences constrained the number of
U.S. forces to be deployed and lives to be lost far more than the amount of
money to be spent and ordnance to be consumed. In response, he and his subordinates instructed the military to ªght what they themselves acknowledged
to be an ineffective, capital- and ªrepower-intensive strategy. The article explains this seemingly nonstrategic behavior using a theory, generalizable beyond this speciªc case, of the distribution of the costs of war within the
electorate.138
Israel’s experience in its 2006 war against Hezbollah suggests that this
phenomenon is not limited to the Vietnam War or to U.S. strategic culture.139
Israel expended 170,000 artillery shells, twice the number ªred in the conven-
proponents (including Komer, Robert Thompson, and Francis “Bing” West) criticized the strategy
of attrition, but only one identiªed signiªcant differences between the Westmoreland and Abrams
approaches.
137. Thayer, War without Fronts, pp. 23–26.
138. This article focuses on democracies, but nondemocracies may also be subject to this pressure.
According to the theory, a state in which the very wealthy controlled policy would pursue a laborintensive doctrine. The more the government must take the average citizen’s preferences into account, the more it will conform to the theory’s predictions.
139. As another example suggesting that the theory is more widely applicable, consider the tremendous public attention paid to the 2007 “surge” of 30,000 U.S. military personnel in Iraq, representing a 20 percent increase in deployed personnel. Less attention has been given to the ªvefold
increase in coalition airstrikes in 2007 over 2006. Anthony H. Cordesman, “Air Combat Trends in
the Afghan and Iraq Wars” (Washington, D.C.: Center for Strategic and International Studies,
2008).
International Security 34:3 156
tional 1973 Arab-Israeli War, in a month.140 The Israel Defense Force’s (IDF’s)
initial campaign plan—a rapid air and small-unit ground assault that relied on
ªrepower to control territory—was designed to minimize the number of
ground forces and casualties. The Israeli cabinet rejected it; the transportation
minister objected to “exposing 40,000 troops to the Lebanese reality.”141 Four
days into the conºict, the IDF deputy chief of staff recommended stopping the
campaign: “We have exhausted the [aerial] effort; we have reached the peak;
from now on we can only descend.”142 Nonetheless, despite its intention
to avoid a ground war, the Israeli government announced ambitious goals far
beyond releasing hostages and deterring further rocket attacks.143 A report
written by a subsequent government commission describing the strategic conundrum evokes the constraints faced by Johnson in Vietnam: “Declared goals
were too ambitious, and it was publicly stated that ªghting will continue till
they are achieved. But the authorized military operations did not enable their
achievement.” The report acknowledges the government’s bind: no “other effective military response to such missile attacks than an extensive and prolonged ground operation” existed, but this “would have a high ‘cost’ and did
not enjoy broad support.”144
Cost distribution theory makes such behavior explicable. From four major assumptions—security is a public good; voters weigh security beneªts
against taxes, conscription, and casualties; median wealth is less than mean
wealth in every state; and the preferences of the median voter are heeded in a
democracy—I derive a voter preference for a capitalized military. Like
the democratic exceptionalist research program, this article ªnds evidence
that the American public weighs the beneªts of limited war against the costs.
Although one recent study of American public opinion assigns expectations of
success as the most important factor in the public’s support for a conºict, it
also points out that the public more generally carries out relatively competent
cost-beneªt analysis.145 This article extends this logic by arguing that when the
ability to substitute matériel for personnel is low, as it is against unconventional opponents, democracies may still prosecute wars using an ill-suited military doctrine (and thus a lower chance of success), because the costs remain
modest for this pivotal voter.
Democratic exceptionalism’s cost-internalization mechanism provides an
140. Kober, “The Israel Defense Forces in the Second Lebanon War,” p. 24.
141. Ibid., pp. 10–11.
142. Ibid., p. 4.
143. Ehud Olmert, “Israel Will Not Be Held Hostage—Olmert’s Address to the Knesset,” Jerusalem
Post, July 18, 2006.
144. “The Winograd Report: The Main Findings of the Winograd Partial Report on the Second
Lebanon War,” Haaretz, January 5, 2007, http://www.haaretz.com/hasen/spages/854051.html.
145. Gelpi, Feaver, and Reiºer, Paying the Human Costs of War, p. 20.
The Myth of Military Myopia 157
overly optimistic assessment of democracies’ discretion in how and even when
they ªght small wars. Neither an apolitical public, nor a dysfunctional military
culture, nor a military doctrine divorced from grand strategy causes a ºawed
warªghting strategy. Rather, it results from political leaders’ assessment of the
average voter’s preferences. Although claiming that democracies substitute
capital for labor to reduce the costs of war for voters is not news, tying the pursuit of such a strategy to a rational voter has two novel and important
implications.
First, many observers argue that most wars of the twenty-ªrst century will
be hybrid conºicts involving unconventional opponents; ªnding the root
cause of poor counterinsurgency is an essential task. This article argues that
ªxating on reforming the armed services (or even the civilian tools of foreign
policy) in an effort to improve democratic performance in small wars is its
own form of myopia. My theory gives reason to be skeptical of how much the
U.S. military will be allowed to shift by future administrations and the public
to which they are held accountable. Dysfunctional organizations can eventually learn and adapt. If the public suffers from foolish preconceptions, it may
be dissuaded through public education and the marketplace of ideas. Even
positing a powerful strategic culture underpinning U.S. doctrine suggests that
“it is at least possible that by deconstructing the standard American ‘way’ . . .
some pathways to improved performance may be identiªed.”146 But if a rational, fully informed electorate views such a military doctrine as its best option,
the prospects for change are less clear.
Second, the distribution of costs and beneªts affects not only how a state
should ªght a small war, but whether it enters such conºicts in the ªrst place.
Although this article has focused on one apparent element of nonstrategic
behavior—poor counterinsurgency—linking doctrine to the voter’s costbeneªt analysis holds potential for understanding why democracies select into
these ªghts. The moral hazard resulting from turning war into an exercise in
ªscal rather than social mobilization may encourage the average voter to support an aggressive grand strategy as well as a military doctrine that ªghts
the resulting conºicts ineffectively. Because of the heavily capitalized nature of
its armed forces, the United States is likely to ªght small wars badly, but continue to ªght them all the same. For a democracy’s average voter, building a
military to ªght these wars of choice inefªciently but often is not a bug in the
program but a feature.
146. Gray, Irregular Enemies and the Essence of Strategy, p. 30.
Who Lost Vietnam? Soldiers, Civilians, and U.S. Military Strategy
James McAllister
International Security, Volume 35, Number 3, Winter 2010/11,
pp. 95-123 (Article)
Published by The MIT Press
For additional information about this article
http://muse.jhu.edu/journals/ins/summary/v035/35.3.mcallister.html
Access Provided by University of Arizona at 08/11/11 5:10PM GMT
Who Lost Vietnam?
Who Lost Vietnam?
James McAllister
Soldiers, Civilians, and
U.S. Military Strategy
W
hy did Gen. William
Westmoreland and the U.S. Army pursue a ºawed military strategy during the
Vietnam War? As the United States ªnds itself involved in a dire counterinsurgency campaign in Afghanistan, with little apparent reason for optimism,
some political scientists have turned their attention to the origins and development of U.S. military strategy in Vietnam for readily applicable lessons. Yet,
despite the passage of four decades and the declassiªcation of millions of
pages of documents, diplomatic and military historians still have not reached a
consensus as to why the United States pursued a ºawed military strategy in
Vietnam. Indeed, historians continue to argue about which elements of the
strategy were ºawed and whether any strategy could have led to success in
Vietnam. For these reasons, political scientists and historians should pay careful attention to a theoretically innovative article by Jonathan Caverley,
published in a recent issue of this journal.1
To appreciate the signiªcance of Caverley’s argument, it is crucial to understand the two dominant schools of thought on U.S. military strategy in
Vietnam. The ªrst school, which George Herring has labeled the “counterinsurgency school,” is best represented by the work of scholars such as Andrew
Krepinevich, Guenter Lewy, Lewis Sorley, and John Nagl.2 In his 1986 book,
The Army and Vietnam, Krepinevich argues that military strategy in Vietnam
was hopelessly ºawed because Westmoreland and the U.S. Army were wedJames McAllister is Professor of Political Science at Williams College and a member of the U.S. State Department’s Historical Advisory Committee on Diplomatic Documentation.
The author would like to thank Richard Betts, Andrew Birtle, Peter Brush, Gian Gentile, Brendan
Green, Robert Jervis, David Kaiser, Andrew Krepinevich, Mark Lawrence, Doug Macdonald, Paul
MacDonald, James Mahon, Mark Moyar, Joshua Rovner, Jerome Slater, James Willbanks, the anonymous reviewers, and the wonderful students in the political science 382 research seminar at Williams College for their insightful comments and suggestions. Finally, he would also like to
acknowledge the invaluable support provided by the Stanley Kaplan Program in American Foreign Policy at Williams College.
1. Jonathan D. Caverley, “The Myth of Military Myopia: Democracy, Small Wars, and Vietnam,”
International Security, Vol. 34, No. 3 (Winter 2009/10), pp. 119–157.
2. George C. Herring, “American Strategy in Vietnam: The Postwar Debate,” Military Affairs, Vol.
46, No. 2 (April 1982), pp. 57–63; Andrew F. Krepinevich Jr., The Army and Vietnam (Baltimore, Md.:
Johns Hopkins University Press, 1986); Guenter Lewy, America in Vietnam (New York: Oxford University Press, 1978); Lewis Sorley, A Better War: The Unexamined Victories and Final Tragedy of America’s Last Years in Vietnam (New York: Harcourt Brace, 1999); and John A. Nagl, Learning to Eat Soup
with a Knife: Counterinsurgency Lessons from Malaya and Vietnam (Chicago: University of Chicago
Press, 2005).
International Security, Vol. 35, No. 3 (Winter 2010/11), pp. 95–123
© 2010 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
95
International Security 35:3 96
ded to a concept of warfare that heavily emphasized overwhelming ªrepower,
search and destroy missions, and conventional uses of force rather than classic
principles of counterinsurgency (COIN) warfare. In Krepinevich’s view, “The
Army Concept of war is, basically, the Army’s perception of how wars ought to
be waged and is reºected in the way the Army organizes its troops for battle.
The characteristics of the Army Concept are two: a focus on mid-intensity, or
conventional, war and a reliance on high volumes of ªrepower to minimize
casualties—in effect, the substitution of material costs at every available opportunity to avoid payment in blood.”3 For Krepinevich and other scholars in
the counterinsurgency school, Westmoreland and the U.S. Army fought ineffectively because they applied a concept of warfare unsuited to their
adversary.
The role of President Lyndon Johnson and his civilian advisers is fairly peripheral in Krepinevich’s account of the origins and evolution of U.S. strategy
on the ground in Vietnam. The Army Concept obviously predated the war,
and Krepinevich argues that the Johnson administration was inclined to let the
army run the ground war as long as it did not expand operations beyond
the borders of South Vietnam.4 Instead of the relentless focus on ªrepower and
search and destroy tactics, Krepinevich and other scholars in this tradition
suggest that Westmoreland and the U.S. Army should have adopted a counterinsurgency strategy based on the model of the Combined Action Platoon
(CAP) program favored by the U.S. Marines in I Corps.
In contrast, the revisionist school, represented by historians such as
John Carland, Dale Andrade, Andrew Birtle, and Mark Moyar, defend
Westmoreland’s conduct of the war against the critical assessments of the
counterinsurgency school.5 The main defense of Westmoreland put forward by
the revisionists is that realities on the ground in South Vietnam—the threat
posed by North Vietnamese and Vietcong (VC) main force units—left him little
3. Krepinevich, The Army and Vietnam, p. 5 (emphasis in original). The essence of Caverley’s “cost
distribution theory” is a reframing of the Army Concept of War: both his theory and Krepinevich’s
concept posit that the United States substituted technology and ªrepower to minimize U.S. casualties in Vietnam.
4. Ibid., p. 165.
5. See John M. Carland, “Winning the Vietnam War: Westmoreland’s Approach in Two Documents,” Journal of Military History, Vol. 68, No. 2 (April 2004), pp. 533–574; Dale Andrade,
“Westmoreland Was Right: Learning the Wrong Lessons from the Vietnam War,” Small Wars & Insurgencies, Vol. 19, No. 2 (June 2008), pp. 145–181; Andrew J. Birtle, “PROVN, Westmoreland, and
the Historians: A Reappraisal,” Journal of Military History, Vol. 72, No. 4 (October 2008), pp. 1213–
1247; and Mark Moyar, Triumph Forsaken: The Vietnam War, 1954–1965 (Cambridge: Cambridge
University Press, 2006), pp. 335–336. Another inºuential critique of U.S. strategy is Harry G. Summers Jr., On Strategy: A Critical Analysis of the Vietnam War (Novato, Calif.: Presidio, 1982). Summers’s critique is revisionist, but his focus is considerably different from those noted above.
Who Lost Vietnam? 97
alternative but to focus his attention and resources on conventional methods of
warfare. This defense is essentially the same as that offered by Westmoreland
in his memoirs: in Vietnam the U.S. Army faced a threat from the main force
units, “the bully boys with crowbars,” and “the termites,” or the guerrilla
forces in the villages. Vietnam was neither a standard conventional struggle
nor a standard insurgency; it was both at the same time. Westmoreland had no
doubt that the army’s focus had to be directed at the bully boys, because they
presented the greatest and most immediate threat.6 In the view of revisionist
historians, Westmoreland and the U.S. Army could not ªght in strict accordance with the principles of counterinsurgency because the threat in South
Vietnam was vastly different and more complex than that posed by lightly
armed and isolated insurgents in villages.
The main lines of division between the counterinsurgency school and the revisionists have been fairly established for decades. Caverley’s argument is important because it presents an alternative explanation for U.S. military strategy
that is at odds with the scholarship of both schools. His main thesis is that democracies such as the United States inevitably do a poor job of combating insurgencies because civilian ofªcials “shift the burden of providing for the
nation’s defense onto the rich by employing capital as a substitute for military
labor.” Caverley argues that for electoral reasons, leaders of democracies are
more willing to expend “treasure rather than blood” because average voters
care more about the latter than the former. As a result, democracies ªght insurgencies poorly because they rely on overwhelming ªrepower and technology
rather than on a more effective COIN strategy requiring large numbers of soldiers.7 The policy implication of Caverley’s thesis and his historical interpretation of Vietnam could not be more depressing for those hoping to either stay
out of or succeed in counterinsurgency conºicts: democracies are likely to become involved in such conºicts, but their leaders will be unable or unwilling
to provide the military with the resources needed to win these wars. For
Caverley, the recent efforts of scholars and military practitioners to develop
successful counterinsurgency practices, symbolized by the U.S. Army/Marine
6. See William C. Westmoreland, A Soldier Reports (Garden City, N.Y.: Doubleday, 1976), p. 145;
and William Westmoreland, “A Military War of Attrition,” in W. Scott Thompson and Donaldson
D. Frizzell, The Lessons of Vietnam (New York: Crane, Russak and Company, 1976), pp. 57–71. For
other accounts of Westmoreland’s strategic thought by his contemporaries, see Phillip B.
Davidson, Vietnam at War: The History, 1946–1975 (New York: Oxford University Press, 1991); Gen.
Bruce Palmer Jr., The 25-Year War: America’s Military Role in Vietnam (Lexington: University of Kentucky Press, 1984); and Adm. U.S.G. Sharp, Strategy for Defeat: Vietnam in Retrospect (Novato, Calif:
Presidio, 1978).
7. Caverley, “The Myth of Military Myopia,” p. 121.
International Security 35:3 98
Corps Counterinsurgency Field Manual, are ultimately futile given the United
States’ democratic political system.8
Historians should examine Caverley’s thesis closely because the author does
not claim only to have made a theoretical contribution with important policy
implications. Caverley also argues that his thesis “challenges the dominant explanation for ºawed counterinsurgency in the Vietnam War, which blames a
myopic military bureaucracy and culture for failing to adapt its conventional
doctrine to the conditions in Vietnam.”9 He claims that “contrary to much of
the conºict’s historiography, Lyndon Johnson’s administration played a crucial
role in rejecting a more labor-intensive COIN approach in favor of a capitalintensive strategy that it understood to be less effective but reºective of the
cost preferences of the average voter.”10
In the above quotations, diplomatic and military historians will recognize
that Caverley is advancing a highly original interpretation of both U.S. military strategy and civil-military relations during the Vietnam War. Two important elements make his argument an outlier to the historical debate
over Vietnam strategy. First, Caverley suggests that the sources of U.S. military
strategy were almost entirely civilian: President Johnson and his civilian
advisers—not Westmoreland and the army—purportedly insisted that the military pursue a conventional strategy based on search and destroy tactics, a focus on the threat of Vietcong main force units, and a reliance on overwhelming
ªrepower. Second, Caverley argues that Westmoreland and the army—not
Johnson and his civilian advisers—wanted to ªght “a better war,” one more
consistent with the classic principles of counterinsurgency warfare. Both during and since the war, Westmoreland has been roundly castigated for his alleged inattention to the importance of paciªcation and counterinsurgency. In
contrast, Caverley portrays Westmoreland as a proponent of greater attention
to paciªcation who was ostensibly not allowed by the Johnson administration to provide the army with the manpower resources necessary to adopt a
paciªcation-oriented strategy.11
8. The U.S Army/Marine Corps Counterinsurgency Field Manual (Chicago: University of Chicago
Press, 2005).
9. Caverley, “The Myth of Military Myopia,” p. 121.
10. Ibid. Caverley seems to assume that the counterinsurgency school is the dominant perspective
among scholars. Revisionists certainly do not blame a “myopic military bureaucracy” for Vietnam.
11. The phrase “a better war” refers to Sorley’s inºuential book about Gen. Creighton Abrams.
Caverley does not explicitly use this phrase, but Sorley’s argument provides an easy way to understand the essence of his argument about civil-military disagreements during the Johnson administration. Where Sorley argued that Abrams fought a “better war” than Westmoreland because
he paid greater attention to paciªcation, Caverely’s implicit argument is that Westmoreland
wanted to ªght a better war than Lyndon Johnson and his top civilian strategists permitted.
Who Lost Vietnam? 99
Caverley’s historical argument about U.S. military strategy is unique
because it does not ªt comfortably into either the counterinsurgency or the revisionist school. The crucial difference between Caverley and the counterinsurgency school is the source of the U.S. preference for a conventional or
capital-intensive military strategy in Vietnam. Where Krepinevich and other
scholars in this school contend that U.S. strategy was rooted in the intellectual
preferences of Westmoreland and the army leadership, Caverley claims that it
ºowed directly from the preferences of Johnson and his civilian advisers.
Caverley’s relationship to the scholarship of the revisionist school is equally
problematic. Scholars in this tradition would undoubtedly welcome his criticisms of the counterinsurgency school, his evident hostility toward the decisions of Johnson and his civilian advisers, as well as his general sympathy for
Westmoreland and the U.S. Army. Yet, they would have tremendous difªculty
accepting the substance of Caverley’s historical interpretation. Revisionist
scholars do not argue that Westmoreland pursued a particular military strategy within the borders of South Vietnam because he was explicitly directed to
do so by the president and his key civilian strategists.12 Although they believe
that Westmoreland was much more sensitive to questions of paciªcation than
often portrayed, they would emphatically reject Caverley’s argument that the
army “focused disproportionately on ªnding, ªxing, and destroying conventional army units” and that paciªcation “was arguably the more important
task . . . and yet was given short shrift.”13
This article poses the following question: Does the historical evidence support an argument that, if accurate, would overturn a massive body of scholarship on the military dimensions of the Vietnam War? I conclude that
the evidence Caverley presents fails to support his interpretation of why the
United States fought as it did in Vietnam. Generating deductive theories and
counterintuitive arguments is indeed a useful endeavor, one as important as
the historian’s goal of understanding the past on the basis of a rigorous examination of primary and secondary sources. At the same time, in seeking to validate their arguments, political scientists must be careful not to neglect or abuse
the historical record. Political scientists must submit to the same rigorous stan12. The caveat “within the borders of South Vietnam” is important because Johnson did prevent
the military from carrying out invasions of Laos, Cambodia, and North Vietnam. Caverley’s thesis
is not about this kind of explicit direction; his article is focused on U.S. ground strategy within the
borders of South Vietnam.
13. Caverley, “The Myth of Military Myopia,” p. 130. Caverley argues, “Consensus exists on the
reasons why the choice of strategies contributed to failure. . . . [T]he United States pursued a
skewed balance of effort between ªghting the enemy’s main and insurgent forces” (ibid.). None of
the revisionist scholars identiªed in footnote 5 would accept the validity of this consensus.
International Security 35:3 100
dards as historians when investigating whether the past conforms to the elegant and deductive theories they have developed. In his article, Caverley often
treats the primary documents and relevant secondary sources in a way that all
serious historians and political scientists should avoid. Indeed, they would
ªnd his account of U.S. military strategy and civil-military relations during the
Vietnam War unrecognizable.
Below I present three counterarguments to Caverley’s thesis. First, I demonstrate that regardless of its merits, Westmoreland pursued his own strategic vision and was not directed to ªght in the manner he did within the borders of
South Vietnam by Johnson or his top civilian advisers. Second, there was simply no great dispute between civilians and the military over paciªcation strategy during the Vietnam War; it is a myth that Westmoreland would have
adopted a COIN strategy if he did not have to follow the preferences of the civilian leadership. Finally, Caverley’s theoretical emphasis on public opinion
and the preferences of the average American voter as the key to understanding
the strategy of the Johnson administration is either unsupported or contradicted by the evidence.
“Instructing” the U.S. Military in Vietnam
The United States waged a capital- and ªrepower-intensive war throughout
Johnson’s term in ofªce. No serious historian could disagree. The crucial issue
that distinguishes Caverley’s thesis from the scholarship of Vietnam War historians is how to account for the adoption of this feature of U.S. military strategy.
Revisionists and scholars of the counterinsurgency school argue that this strategy can be traced either to the intellectual and organizational preferences of
the United States military or to the realities faced by the army on the ground in
Vietnam. Caverley’s thesis presents a challenge to both schools because he argues that civilian preferences account for the nature of U.S. military strategy.
He argues that testing his theory “requires empirical support” for the proposition that “policymakers should understand the hazards of applying such an
approach [waging a capital-intensive campaign] and recommend it anyway. . . . [G]overnment ofªcials must explicitly direct the military to ªght accordingly.” In the concluding section, Caverley claims that his article has
shown that President Johnson and his subordinates “instructed the military
to ªght what they themselves acknowledged to be an ineffective, capital- and
ªrepower-intensive strategy.”14
14. Ibid., pp. 131, 155.
Who Lost Vietnam? 101
Empirical support for these claims would constitute a revolutionary contribution to the historical literature. Historians of both schools are very critical of
Johnson’s leadership during the Vietnam War, but they do not argue that the
president and his civilian advisers advocated a military strategy knowing that
it would be ineffective and that a better strategy was readily available. More
important, although many historians are highly critical about civilian interference in some of the military dimensions of the Vietnam War, scholars do not
argue that Johnson and his civilian advisers “explicitly directed” or “instructed” the military on how to ªght within the borders of South Vietnam. An
examination of the evidence about ªrepower, airpower, and the ground war
suggests that Caverley’s challenge to the historical consensus is unpersuasive.
ªrepower in vietnam
Historians have richly documented the lavish and often excessive use of
ªrepower in Vietnam. It is a truism for both revisionists and scholars in the
counterinsurgency school that the U.S. military preferred to expend treasure
rather than blood in Vietnam. As one infantryman in Vietnam said, “When
you are trying to survive, there is no such thing as too much ªrepower. We
were not military tacticians or logistics people, we were infantrymen trying to
kill them before they killed us.”15 These understandable sentiments were present throughout the hierarchy of the U.S. Army. That lavish ªrepower should
be employed in Vietnam to save American lives constituted the essence of the
tactics recommended by Gen. William E. DePuy, one of Westmoreland’s most
trusted subordinates. Westmoreland’s ideas about ªrepower were the same
as DePuy’s. In addition to saving American lives, however, Westmoreland and
DePuy thought that the lavish use of ªrepower was the best way to combat the
insurgency in Vietnam. As Krepinevich notes, “When General Westmoreland
was asked at a press conference what the answer to the insurgency was, his reply was one word: ‘ªrepower.’”16
Civilian ofªcials may have shared these sentiments, but there is no evidence
that they did or needed to instruct Westmoreland or DePuy to ªght accordingly in Vietnam. Caverley is not explicit about the source of the United States’
enthusiasm for lavish ªrepower, but he does imply that civilians led the way
and the military followed. For example, he argues that “[a]dministration of15. Quoted in Eric M. Bergerud, The Dynamics of Defeat: The Vietnam War in Hau Nghia Province
(Boulder, Colo.: Westview, 1991), p. 175.
16. Krepinevich, The Army and Vietnam, p. 197. For DePuy’s thoughts on the role of ªrepower in
Vietnam, see Henry G. Cole, General William E. DePuy: Preparing the Army for Modern War
(Lexington: University of Kentucky Press, 2008), pp. 143–198.
International Security 35:3 102
ªcials set daily requirements for ordnance expenditure and air strikes,” which
he claims led to complaints from Westmoreland.17 The implication of this statement is that the Johnson administration set out requirements that the military
had no choice but to meet, regardless of their utility. If true, this would constitute a new and crucially signiªcant argument about military strategy in Vietnam, one that has escaped the attention of historians who have spent decades
studying the subject. Yet, Caverley’s evidence for this crucial claim about daily
requirements for ordnance expenditure and air strikes is nothing more than a
single ambiguous sentence from a single July 1965 document.18 Scholars who
have extensively studied ordnance expenditure in Vietnam provide no support for the argument that civilian ofªcials played any role in setting daily
requirements for the expenditure of ªrepower. For example, in the very important case of harassment and interdiction ªre (H&I), a major source of ordnance
expenditure in Vietnam, there is no evidence to support the idea that civilians
played any role in the army’s decisions to rely on H&I ªre in Vietnam.19
airpower in vietnam
Like ªrepower, the U.S. military certainly did not need to be instructed that
airpower should be a central component of the ªghting in Vietnam. Indeed,
the extensive historical literature shows that the main source of conºict
was the desire of the Johnson administration to wage the air war more gradually and to impose some restrictions on the military.20 As George Herring ar17. Caverley, “The Myth of Military Myopia,” p. 142.
18. The only document Caverley cites to back up this claim concerning daily requirements for
ordnance expenditure and air strikes is a July 17, 1965, cable from Westmoreland to Admiral
Sharp. The only possibly relevant sentence in the memo is the following: “During MACV brieªng
yesterday afternoon, Secretary McNamara manifested uncommon interest in our disclosure that
small percentage of daily requirements for tactical airstrikes not always fulªlled by available
USAF [U.S. Air Force] in-country assets and allocated CVA [aircraft carrier] resources.” It is unclear what this document says of any signiªcance to Caverley’s claim.
19. See John M. Hawkins, “The Costs of Artillery: Eliminating Harassment and Interdiction Fire
during the Vietnam War,” Journal of Military History, Vol. 70, No. 1 (January 2006), pp. 91–122. A
similar example is the case of close air support. Caverley cites a single document from December
1965 to argue, “McNamara was constantly pushing Wheeler and Westmoreland to improve the
close air support provided to army ground forces.” Caverley, “The Myth of Military Myopia,”
p. 142. Even if the document lends itself to that interpretation, it is hard to see how or why
Wheeler and Westmoreland would be opposed to McNamara’s efforts, and there is no evidence
cited by Caverley that the military was opposed to improving close air support. The lengthy and
deªnitive ofªcial study of close air support does not even hint at a conºict between McNamara
and the army over the provision of close air support in Vietnam. See John Schlight, Help from
Above: Air Force Close Air Support of the Army, 1946–1973 (Washington, D.C.: Air Force History and
Museums Program, 2003).
20. Adequately summarizing the massive literature on airpower in Vietnam is not possible to do
here. See, for example, Robert A. Pape, Bombing to Win: Air Power and Coercion in War (Ithaca, N.Y.:
Who Lost Vietnam? 103
gues, “The military in general pressed for a quick, hard-hitting bombing
campaign against major North Vietnamese targets, although the navy and air
force differed sharply on the focus and major objectives of the bombing. The
civilians and especially President Johnson sought to keep the bombing under
tight restriction.”21
An example of how Caverley incorrectly suggests that civilians made
decisions about airpower without military input is his discussion of the use of
B-52s in Vietnam. He argues that crucial decisions such as using B-52s to bomb
Vietcong sanctuaries were reached by the Johnson administration “with little
strategic input from the JCS [Joint Chiefs of Staff].”22 Yet, the Johnson administration reached this decision with extensive military input: the JCS recommended the use of B-52s against North Vietnam as early as November 1964,
but this advice was turned down by the president. The driving force behind
the use of B-52s in South Vietnam was not Johnson or his advisers, but
Westmoreland. The general made a strong argument to use B-52s to attack
Vietcong bases in May 1965, and both the JCS and the commander in chief of
the U.S. Paciªc Command approved their use before the ªrst “Arc Light”
strikes of June 1965.23 In his memoirs, Westmoreland expresses enthusiasm for
the use of B-52s in South Vietnam and displeasure with civilian authorities
who were not always willing to approve their use.24 Civilians may have supported the use of B-52s, and no historian would argue otherwise, but they did
not need to instruct the military to employ them in Vietnam.
Cornell University Press, 1996), pp. 174–210; and Mark Clodfelter, The Limits of Airpower: The American Bombing of North Vietnam (Lincoln: University of Nebraska Press, 1989). For a representative
view of the military’s critique of civilian interference in the air war, see Sharp, Strategy for Defeat.
Talking about this issue in terms of civilians and the military ignores the real debates that took
place within these groups. My point is simply that the U.S. military needed no encouragement or
direction in waging a capital-intensive war in Vietnam and that military frustration with civilians
was caused by their perception of gradualism and limitations rather than by any push by Johnson
and civilians to employ more airpower than the military believed was warranted.
21. George C. Herring, LBJ and Vietnam: A Different Kind of War (Austin: University of Texas Press,
1994), p. 43.
22. Caverley, “The Myth of Military Myopia,” p. 149. Caverley cites the work of H.R. McMaster to
make this point about JCS exclusion. It is true that Johnson never relied on the entire JCS to the extent McMaster would have preferred. McMaster’s book, however, emphasizes that the JCS wanted
to go much further than Johnson did when it came to employing airpower. Greater strategic input
from the JCS would have resulted in more, rather than less, employment of B-52s in Vietnam. See
McMaster, Dereliction of Duty: Lyndon Johnson, Robert McNamara, the Joint Chiefs of Staff, and the Lies
That Led to Vietnam (New York: HarperCollins, 1997), pp. 64–79.
23. On the use of B-52s in Vietnam, see Graham A. Cosmas, MACV: The Joint Command in the Years
of Escalation, 1962–1967 (Washington, D.C.: Center of Military History, 2006), pp. 325–327; and John
T. Correll, “Arc Light,” Air Force Magazine, Vol. 92, No. 1 (January 2009), pp. 58–62.
24. See Westmoreland, A Soldier Reports, p. 137.
International Security 35:3 104
the ground war in vietnam
Similar problems appear in Caverley’s discussion of the ground war in South
Vietnam. Johnson did explicitly instruct Westmoreland not to invade Laos,
Cambodia, or North Vietnam. There is no evidence, however, to suggest that
the president and his civilian advisers directed Westmoreland on how he
should ªght within the borders of South Vietnam. Although Westmoreland
would later criticize the Johnson administration for not allowing him to invade
Vietcong sanctuaries and for imposing restrictions on the air campaign,
he never argued that Johnson or Secretary of Defense Robert McNamara
instructed him to ªght on the ground in the way that he did within South
Vietnam. Rather, as he writes in his memoirs, “Once the penchant of niggling
ofªcials in Washington for quibbling over B-52 bomb targets had passed,
President Johnson and Secretary McNamara afforded me marked independence in how I ran the war within the borders of South Vietnam.”25 As John
Carland notes in his ofªcial history of the U.S. Army in Vietnam, “The conºict
in South Vietnam in 1965 and 1966 was truly Westmoreland’s war.”26
A much weaker variant of Caverley’s argument might be that although the
administration did not explicitly direct Westmoreland how to ªght in Vietnam,
manpower limitations forced him to rely on excessive ªrepower and to forego
policies more in line with counterinsurgency principles. Indeed, Caverley
makes this argument regarding the use of tanks and the failure of the United
States to expand the CAP program throughout South Vietnam. On the subject
of tanks, Caverley states that “the generals understood that heavy forces were
inappropriate for the terrain and the type of war being fought.” He provides
two quotations from Westmoreland and Army Chief of Staff Gen. Harold
Johnson that reºect this position in the summer of 1965. Why were tanks eventually employed extensively in Vietnam if the top two ofªcials in the army
thought they were inappropriate? Caverley offers a simple and seemingly
plausible argument, one consistent with his larger theoretical argument about
capital as a substitute for manpower: “The army’s shift to armor occurred
25. Ibid., p. 261. For Westmoreland’s view that he had “a carte blanche” and independence to run
the war in South Vietnam as he liked, see his revealing oral history remarks, quoted in John M.
Carland, Combat Operations: Stemming the Tide: May 1965–October 1966 (Washington, D.C.: Government Printing Ofªce, 2000), p. 363.
26. Carland, Stemming the Tide, p. 363. Caverley’s own argument basically concedes Westmoreland’s independence in running the ground war when he says that “the commander in the ªeld
was doing precisely what the president wanted him to do.” Caverley, “The Myth of Military Myopia,” p. 145. There are two problems with this formulation. First, it is impossible to reconcile with
Caverley’s own standard for testing his theory given that there is no explicit direction from civilians. Second, it concedes a central point of the counterinsurgency and revisionist schools, namely,
that Westmoreland operated according to his own strategic conception in Vietnam.
Who Lost Vietnam? 105
gradually over 1966 only as the need to substitute for labor became more
apparent.”27
There is no controversy in the historical literature regarding why tanks came
to be employed by the U.S. Army in Vietnam. Caverley cites three primary
documents to support his argument about tanks, but to assess his claim, one
need only look at the single page of text from the lone secondary source he
cites: Gen. Donn Starry’s Armored Combat in Vietnam.28 According to Starry, in
the summer of 1965, both Westmoreland and Johnson made exceptions for the
use of tanks in Vietnam. Westmoreland conceded that tanks and mechanized
infantry units might be suitable in “a few coastal areas, most notably in the I
Corps area.” Johnson was willing to allow the 1st Squadron, Fourth Cavalry to
retain its medium tanks and to potentially reinforce them later.29 The page
from Starry’s book that Caverley cites, however, fails to support the argument
that the greater use of tanks was a response to insufªcient manpower. Starry’s
explanation for the increased use of armor in Vietnam was simple: combat experience and a comprehensive army study in 1967 proved that skeptics of
tanks were wrong. Westmoreland himself acknowledged during the war and
in his memoirs that his initial views on tanks were incorrect: “The ability of
mechanized cavalry to operate effectively in the Vietnamese countryside convinced me that I was mistaken in a belief that modern armor had only a limited role in the ªghting in Vietnam. . . . [M]uch of the land was in fact solid
terrain, and roads and bridges could be improved to accommodate tanks.”30
Caverley appears to have posited an explanation consistent with his
theory—tanks as a substitute for manpower—without any historical evidence
to support it. This substitution of theory for historical evidence also drives his
analysis of the CAP program. The program was by far the best-known counterinsurgency alternative to Westmoreland’s strategy of attrition, and numerous scholars have put it forward to rebut his claim that no other viable ground
strategy existed in Vietnam. The CAPs, which consisted of small numbers of
lightly armed U.S. Marines living and ªghting with South Vietnamese forces
in villages, ªt the deªnition of a noncapital-intensive method of counterinsurgency. Whether they required more or less manpower, and whether their employment on a larger scale would have led to more or fewer U.S. casualties, are
still issues of historical debate. Krepinevich and other scholars of the counter-
27.
28.
29.
30.
Caverley, “The Myth of Military Myopia,” p. 146.
Donn A. Starry, Armored Combat in Vietnam (New York: Arno, 1980).
Ibid., p. 56.
Westmoreland, A Soldier Reports, pp. 178, 415, at p. 178.
International Security 35:3 106
insurgency school argue that the CAP program could have protected the population more effectively and with far fewer troops than Westmoreland was
employing in search and destroy operations.31 The revisionist school remains
skeptical about both the merits and wider viability of the CAP program.
Scholars in this tradition do suggest that an adoption of the program by the
army would have required more troops than Westmoreland had at his
disposal.32
Caverley’s argument is roughly consistent with the revisionist view, but he
overstates the role of civilians and minimizes Westmoreland’s reasons for opposing the expansion of the CAP program throughout South Vietnam. The
header for this discussion in his article is “Civilian Rejection of the Marines’
Combined Action Program,” and he ends with a statement that although the
CAP program was potentially a viable alternative strategy, “the civilian leadership was unwilling to take that chance.”33 There is, however, virtually no evidence that Johnson or his civilian advisers ever debated or seriously
considered the CAP question. The best evidence that a high-ranking civilian
was opposed to the expansion of the CAP concept can be found in Victor
Krulak’s memoirs, First to Fight: An Inside View of the U.S. Marine Corps. Krulak
writes that he tried to convince McNamara of the merits of the Marine
paciªcation concepts, but that the secretary of defense thought they would require too much time to work. Krulak emphasizes, however, that the main obstacle to the widespread adoption of the CAP program was Westmoreland’s
opposition: “Sound and logical as it appeared, the Marines’ strategy had two
defects: General Westmoreland did not agree with it; and it was unable to address the reality that the enemy enjoyed a privileged sanctuary in the ports of
North Vietnam and in Laos, through which a cascade of deadly munitions was
ºowing.”34
31. Krepinevich, The Army and Vietnam, pp. 171–177.
32. The best revisionist analysis of the army critique of the CAP program can be found in Andrew
Birtle’s U.S. Army Counterinsurgency and Contingency Operations Doctrine, 1942–1976 (Washington,
D.C.: Center of Military History, 2006), pp. 399–400. Revisionists argue that Westmoreland did not
have the troops because infantrymen were only a small fraction of U.S. Army personnel. The counterinsurgency school rejects this argument because, as Krepinevich argues, the army’s “force mix
was grossly overweighted in favor of support personnel necessary to maintain the ªre-power intensive U.S. forces deployed in South Vietnam.” Krepinevich, The Army and Vietnam, p. 176.
33. Caverley, “The Myth of Military Myopia,” p. 153.
34. Victor H. Krulak, First to Fight: An Inside View of the U.S. Marine Corps (Annapolis: U.S. Naval
lnstitute, 1984), pp. 194, 204. Krulak was a lieutenant general in the U.S. Marine Corps and a critic
of Westmoreland’s approach to the war. His memoirs illustrate another serious problem with
Caverley’s theoretical framework: some of the most fervent military advocates of a COIN strategy
were also the most enthusiastic advocates of a capital-intensive strategy. For example, Krulak also
told McNamara that the United States needed “constant bombardment from the sea of every sen-
Who Lost Vietnam? 107
Caverley supports his argument about the CAP program with a quotation
from Westmoreland: “I simply had not enough numbers to put a squad of
Americans in every village and hamlet.” The omitted part of the sentence,
however, is important: “[T]hat would have been fragmenting resources and
exposing them to defeat in detail.”35 The full quotation makes clear that
Westmoreland based his opposition to an expansion of the CAP program on
more than limited manpower. Earlier in his memoirs, Westmoreland was even
more explicit about his general aversion to dispersing small numbers of U.S.
soldiers throughout the villages of South Vietnam. In his view, such a strategy
ignored “the undisputed military canon that to try to be strong everywhere is
to be weak everywhere.”36 As Mark Moyar and Stephen Rosen point out,
Westmoreland’s dislike of breaking big units into smaller units was heavily
inºuenced by his advice to South Vietnamese commanders in Binh Dinh in
1964 to break their forces into smaller units, a decision that led to serious
defeats.37
The Johnson administration did not allow the U.S. military to run the war
exactly as Westmoreland or the JCS would have liked. Few if any historians
would argue they did. This historical consensus, however, is fundamentally
different from Caverley’s argument that the Johnson administration explicitly
directed the military on how to ªght in Vietnam. To prove his argument,
Caverley would have to show that the administration explicitly directed the
U.S. military to ªght in a way that went against their own traditions or assessment of what needed to be done in Vietnam. The evidence he presents does
not show that civilians pushed a reluctant military to expend a massive
amount of ªrepower in Vietnam; Westmoreland and DePuy needed no instructions in this area because their own intellectual convictions already led them to
this conclusion. In the case of airpower, if anything, civilians attempted to instruct the military to recognize the merits of gradualism and limitations, an effort that largely failed. As the evidence shows, Westmoreland certainly did not
employ tanks and B-52s as a result of civilian direction or as a consequence of
perceived manpower limitations. The lack of manpower may have made
Westmoreland reluctant to experiment with expanding the CAP program,
but it was just one of many factors in the equation, and it is far from clear
sitive point on the enemy communications, transportation and air defense system within naval
gun range.” Ibid., p. 204.
35. Westmoreland, A Soldier Reports, p. 166.
36. Ibid., p. 147.
37. Moyar, Triumph Forsaken, p. 335; and Stephen Peter Rosen, “Vietnam and the American Theory
of Limited War,” International Security, Vol. 7, No. 2 (Fall 1982), pp. 105–106.
International Security 35:3 108
whether it was the most important one. Caverley provides no compelling reason to reject the historical consensus that primarily emphasizes the central importance of Westmoreland’s opposition to the CAP program.
A Better War? Paciªcation and COIN
Caverley correctly notes that “U.S. forces were for the most part excluded from
paciªcation efforts, leaving these tasks to a South Vietnamese military
and government clearly not competent in these missions.”38 He is wrong, however, in arguing that this exclusion was promoted solely by civilian policymakers against the better advice or recommendations of the military. As
Westmoreland argued in his memoirs, “A cardinal principle of paciªcation
was that it was primarily a South Vietnamese task. The Americans could only
help with advice and resources. The goal was to provide the people with security, social justice, education, medical care, and economic opportunity, and
in the long run only the South Vietnamese could achieve that.”39 Rightly
or wrongly, and many historians would take opposite sides in such a debate,
the historical evidence suggests overwhelmingly that Westmoreland and the
Johnson administration agreed on rejecting a labor-intensive COIN strategy:
U.S. ground forces could take over the main force war, but labor for the “other
war” could and should be provided by the South Vietnamese. This consensus
was based on more than manpower considerations.
U.S. presidents from Harry Truman to Lyndon Johnson never wanted
to send substantial ground forces to South Vietnam. The consistent goal of
policymakers was to create a stable South Vietnam that could preserve its independence without requiring a U.S. military presence. What eventually transformed this commitment of advice and support to what would become the
Vietnam War in 1965 was the stark realization that without an infusion of U.S.
ground forces, the South Vietnamese government would no longer be able to
resist the Vietcong. On June 7, 1965, Westmoreland informed the JCS that he
saw “no course of action open to us except to reinforce our efforts in SVN
[South Vietnam] with additional U.S. or third country forces as rapidly as
practical. . . . I am convinced that U.S. troops with their energy, mobility, and
ªrepower can successfully take the ªght to the VC. The basic purpose of the
additional deployments recommended below is to give us a substantial and
38. Caverley, “The Myth of Military Myopia,” p. 131.
39. Westmoreland, A Soldier Reports, p. 216. Caverley should be commended for his use of primary
documents, but his failure to consult Westmoreland’s memoirs in his article is troubling.
Who Lost Vietnam? 109
hard hitting offensive capability on the ground to convince the VC that they
cannot win.”40 Less than a week later, Westmoreland further outlined his concept of operations for U.S. forces in South Vietnam. The South Vietnamese
faced two distinct but related threats: large, well-equipped military forces and
“the guerrilla, the assassin, the terrorist and the informer.” In Westmoreland’s
view, U.S. ground forces could contribute substantially in dealing with the ªrst
threat, but “only the Vietnamese can make real progress and succeed” with respect to the second problem.41
Westmoreland’s analysis of the situation in South Vietnam framed all of the
discussions that led to the decision to send substantial numbers of U.S. ground
forces in July 1965. The very real differences among civilian and military
ofªcials over the conduct of the war should not obscure the fundamental acceptance of Westmoreland’s concept of operations. The idea that U.S. ground
forces should not be directly involved in the paciªcation effort can be found in
countless documents, and it was a basic principle that was rarely challenged.
Neither the Johnson administration nor Westmoreland ever contemplated
adopting a labor-intensive COIN or paciªcation strategy as deªned by
Caverley. A labor-intensive strategy might have secured the peasants of South
Vietnam from Vietcong predation in the short term, but it would have prevented what was most important for the long-term success of U.S. strategy: the
creation of a political regime in South Vietnam that could progressively pacify
the countryside and deal with the Vietcong over time. No one denied that
paciªcation success was crucial or that the United States needed to assist the
government of South Vietnam in the effort to secure the population, but
the provision of labor was always assumed to be the responsibility of the
South Vietnamese.
Caverley’s arguments about COIN and the Vietnam War stand well
outside the historical consensus on the subject. His main argument is that
Westmoreland and the military repeatedly tried to convince civilian ofªcials to
adopt a labor-intensive COIN strategy and that “the president summarily rejected the COIN option on multiple occasions.”42 He argues that from September 1965 to the summer of 1967 Westmoreland’s efforts to move toward COIN
and paciªcation were repeatedly rebuffed by an administration that wanted to
40. Westmoreland to JCS, June 7, 1965, Foreign Relations of the United States (FRUS), 1964–1968, Vol.
2: Vietnam: January–June 1965, Doc. 337 (Washington, D.C.: Ofªce of the Historian, Bureau of Public
Affairs, U.S. Department of State, 1996), p. 735.
41. Westmoreland to Sharp, June 13, 1965, FRUS, 1964–1968, Vol. 3: Vietnam: June–December 1965,
Doc. 1, p. 2.
42. Caverley, “The Myth of Military Myopia,” p. 146.
International Security 35:3 110
focus on airpower, an anti-inªltration barrier, and the main force war. If accurate, this argument would render entire shelves of books on the Vietnam
War obsolete. Does the evidence that Caverley presents demonstrate that
Westmoreland and other military ofªcials recommended a different paciªcation course that the Johnson administration consistently and “summarily”
rejected?
The evidence overwhelmingly suggests that Westmoreland did not recommend such a course. Caverley cites three examples to argue that the administration rejected a labor-intensive COIN approach. The ªrst is a memorandum
from National Security Adviser McGeorge Bundy to Johnson on September 23,
1965, concerning the subject of what should be done in the event the Vietcong
avoided main force battles and returned to guerrilla tactics. Caverley claims
that in a memo to the president Bundy “summarized Westmoreland’s strategy
before recommending its rejection.”43 According to Bundy, Westmoreland’s
strategy for the war was as follows:
1. Halt the VC offensive.
2. Destroy VC units where they can be found and pacify selected high priority
areas.
3. Restore progressively the entire country to GVN [Government of (South)
Vietnam] control.
4. Support “rural reconstruction” with comprehensive attention to the pacification process.
5. Continue the air campaign against the DRV [North Vietnam] and infiltration routes into Laos.44
Nowhere in this memo, however, does Bundy even remotely suggest rejecting Westmoreland’s strategy. Nor is there evidence that Johnson, who saw the
document, responded to it, which is not surprising given that it contained
nothing new and did not ask the president for a response.45 Westmoreland’s
September 1965 thoughts on the use of U.S. forces in Vietnam were no different
from what they had been three months earlier. In a document from this period
43. Ibid., p. 150.
44. Bundy to Johnson, September 23, 1965, FRUS, 1964–1968, Vol. 3, Doc. 151, p. 414. Caverley did
not include points 1 or 5 in his description of Westmoreland’s concept. By truncating the document, he conveys the impression that Westmoreland put forward a strict paciªcation concept
rather than the actual balanced conception of main force war, airpower, and paciªcation.
45. Westmoreland addressed the civilian inquiries of September 1965 about his strategic concepts
in his memoirs. He was furious that lower-level civilians asked Ambassador Henry Cabot Lodge
to prepare military plans, which he argued were not “valid considerations for the Department of
State.” He does not suggest that there were any substantive conºicts over strategy, and it is clear
that he viewed these inquiries as nothing more than a nuisance. See Westmoreland, A Soldier Reports, p. 161.
Who Lost Vietnam? 111
that John Carland has aptly described as “Westmoreland’s Theory of Victory,”
MACV [Military Assistance Command, Vietnam] strategists reiterated the basic division of labor that would exist between U.S. and South Vietnamese
forces. U.S. forces, and to a lesser extent South Vietnamese forces, would have
primary responsibility for ªnding and clearing out Vietcong main force units.
The securing of territory after the expulsion of main force units was left to the
South Vietnamese Regional and Popular Forces. MACV strategists argued that
U.S. forces would clear while the South Vietnamese would hold: “Except in the
immediate environs of US base areas, US forces are not expected to engage in
securing operations.”46
u.s. forces and paciªcation, 1966
Caverley’s second example of civil-military disagreement over paciªcation revolves around a concept of operations that Westmoreland sent to the JCS on
August 26, 1966. The context of this document is important. Both civilian and
military ofªcials were utterly dismayed by the lack of progress being made in
paciªcation since the introduction of U.S. ground forces a year earlier, with all
agreeing on the primary source of the problem: the poor performance of the
GVN and ARVN [Army of the Republic of (South) Vietnam]. To improve
paciªcation’s chances of success, Johnson appointed Robert Komer to be his
special assistant in running what had come to be called “the other war.”47 Almost a month before Westmoreland submitted his concept of operations, Secretary of State Dean Rusk cabled [Ambassador] Henry Cabot Lodge to say that
Johnson wanted him to “personally address the question of best means of accelerating and giving direction to the paciªcation effort.”48 Westmoreland’s
concept of operations was a reºection of this shared alarm about paciªcation,
but it was hardly new, and it did not call for any signiªcant change in U.S. military strategy. Having reviewed the development of strategy since May 1965,
the general argued that U.S. and “Free World” forces “must take the ªght to
the enemy by attacking his main forces and invading his base areas. . . . Forces
46. MACV, “Tactics and Techniques for Employment of U.S. in the Republic of Vietnam,” September 17, 1965, enclosed in Carland, “Winning the Vietnam War,” p. 557. McNamara met with the
JCS on September 20, 1965, to discuss Westmoreland’s concept. His concept was described in
the following manner with no apparent disagreement: “U.S. forces would prevent main force concentrations; combination of U.S. and RVN [Republic of Vietnam] forces to attack main force sanctuaries that had been in existence for 25 years, move in and drive them out, then police by RVN.”
See JCS Executive Session with Secretary of Defense, September 20, 1965, FRUS, 1964–1968, Vol. 3,
Doc. 145, p. 398.
47. For the best short overview of Komer and U.S. paciªcation programs, see Richard A. Hunt,
Paciªcation: The American Struggle for Vietnam’s Hearts and Minds (Boulder, Colo.: Westview, 1995).
48. Rusk to Lodge, August 3, 1966, FRUS, 1964–1968, Vol. 4: Vietnam: 1966, Doc. 200, p. 555.
International Security 35:3 112
and bases thus discovered will be subjected to either ground attack or quick reaction B-52 and tactical air strikes.” In addition, Westmoreland argued that the
increase in forces at his disposal “will provide the shield that will permit
ARVN to shift its weight of effort to an extent not heretofore feasible to direct
support of revolutionary development. Also, I visualize that a signiªcant number of the US/Free World Maneuver Battalions will be committed to Tactical
Areas of Responsibility (TAOR) missions.”49
Westmoreland was not proposing a “new” concept of operations or a change
in basic military strategy. Offensive operations designed to locate main force
units and their bases remained the focus of his strategy. With the great expansion of ground forces that had taken place since the summer of 1965,
Westmoreland understandably believed that U.S. forces could undertake a
somewhat wider range of tasks. He was not proposing a shift in strategy
toward paciªcation, however, and he was not asking for more ground forces
to take on new tasks; it was ARVN and not the U.S. Army that would
“shift its weight of effort.” Although Caverley is not alone in thinking that
Westmoreland was proposing a new concept, there is little evidence to support
the idea that the general was changing his approach to the war. In fact,
Westmoreland did not put forward his new concept as an issue for debate by
the JCS or civilians; his presentation of the concept was simply intended to be
a summary of his military plans for the upcoming year.50
Caverley argues incorrectly that civilians within the Johnson administration
were hostile to Westmoreland’s concept of operations, even though it came
with a ringing endorsement from Ambassador Lodge. He cites a memo from
Maxwell Taylor expressing concern about the implications of the concept for
U.S. ground forces, which he claims National Security Adviser Walt Rostow
endorsed. He also argues that Johnson ended any further consideration of
Westmoreland’s concept by ordering Rostow to make it clear to the general
that his ideas did not have administration approval. The end of the story allegedly comes in November 1966 when Secretary of State Rusk informs Ambassa49. Rostow to Johnson, August 29, 1966, with enclosed Westmoreland “Concept of Operations,”
FRUS, 1964–1968, Vol. 4, Doc. 220, pp. 603–606, at p. 605. Caverley’s summary of the document
omits the part about ARVN shifting its weight to paciªcation as well as the continuing emphasis
on search and destroy.
50. As Andrew Birtle argues, Westmoreland’s concept of operations in August 1966 was not at all
new and was consistent with ideas he had held all along. If Westmoreland was indeed proposing a
profound transformation in U.S. military strategy in August 1966, it is surely curious that he never
once mentioned this change of heart in his memoirs. While contradicting Caverley on virtually
every other point concerning the civilian reaction to Westmoreland’s concept, John Nagl argues
that the concept did represent a shift in the general’s strategy. See Birtle, “PROVN, Westmoreland,
and the Historians,” pp. 1221–1222; and Nagl, Learning to Eat Soup with a Knife, pp. 160–164.
Who Lost Vietnam? 113
dor Lodge that all of the top ofªcials within the administration, including the
president, believe that “there should be no thought of U.S. taking on substantial share of paciªcation.”51
To the extent there was a new emphasis on paciªcation in Westmoreland’s
concept, however, it was welcomed by civilian ofªcials with enthusiasm rather
than hostility. Caverley does not point out that Rostow sent the document to
Johnson, having written at the top of the document that it was “clearly, in the
right direction; although he and Lodge must engage [Premier Nguyen Cao] Ky
and the ARVN fully if it is to work.” Nor does he mention that the president
was also enthusiastic and wrote on the document: “Let’s get Komer to pick up
and spark this inspiration.”52 Taylor did indeed raise the concerns Caverley
noted, but they were not “endorsed” by Rostow, who merely forwarded the
document to Johnson, noting that Taylor’s concerns could be addressed by getting ARVN more involved in paciªcation. Taylor’s concerns reºected his longstanding reservations about U.S. ground forces in Vietnam.53 The Johnson
administration did not send any message to Westmoreland indicating either
approval or disapproval of his concept of operations, but it did use the opportunity to transform U.S. paciªcation efforts. Komer and McNamara had long
believed that revitalizing paciªcation was largely a management problem, and
their solution was to create an organization that could enable the United States
to better assist the South Vietnamese. This idea, which later became Civil Operations and Revolutionary Development Support, had Johnson’s full backing,
but it was bitterly opposed by Lodge, the State Department, and the Central
Intelligence Agency because they did not want to put control of paciªcation in
the hands of the military.
It is this context, clearly laid out in the Foreign Relations of the United States
volume, that Rusk’s admonition to Lodge in the November 1966 cable needs to
be understood. Lodge had enthusiastically endorsed Westmoreland’s concept
of operations in August, but the two men were certainly not allies in the debate over strategy. While Lodge and Westmoreland maintained cordial relations, they had a fundamental and long-standing philosophical disagreement
about U.S. strategy in Vietnam. Lodge was a staunch critic of MACV’s emphasis on search and destroy missions, which he argued were appropriate for
World War II but not for Vietnam. Lodge did want to use more U.S. forces in
51. Caverley, “The Myth of Military Myopia,” pp. 147, 150–151, at p. 147.
52. Rostow to Johnson, p. 603.
53. By this time, Taylor was merely a special consultant to Johnson and had little power or
inºuence.
International Security 35:3 114
support of paciªcation, and he suggested that these forces would be “10% of
the total men under arms (90% of whom would be Vietnamese) which would
get the whole thing moving.”54 The philosophical disagreement over U.S.
strategy was a minor element in Lodge’s argument with the administration in
November 1966; the important question concerned the management of
paciªcation. Rusk’s cable, with the exception of the last paragraph cited by
Caverley, was devoted entirely to the management issue, and it was Lodge’s
opposition to putting it in the hands of MACV that put him at odds with the
Johnson administration.
Johnson wrote a letter to Lodge four days after Rusk’s cable that accurately
presented the position of his administration. The president made it clear that
neither he nor any of his top civilian and military advisers had a problem with
having a limited number of U.S. forces devoted to paciªcation: “Bob
McNamara and the Joint Chiefs of Staff realize, as does General Westmoreland
on the basis of dispositions that he is already making, that a limited number of
U.S. combat forces must be detailed to be the catalysts for the Vietnamese. . . .
If showing ARVN the way on paciªcation can take up to ten percent of our
troops, it also deserves the full time attention of some of our best generals.”55
In his response to Rusk, Lodge also made it clear that the difference between
his position and that of the administration was minor. As he wrote, “I have always believed that Revolutionary Development/Paciªcation must be carried
out by Vietnamese forces, who, as you say, must establish constructive relations with the population. I have never advocated U.S. forces taking on a “substantial” share of this task.”56 If Westmoreland believed that the administration
was rejecting a new concept of operations in 1966, he kept those thoughts to
himself both at the time and later in his memoirs.
“the philosophy of the war,” 1967
The third example that Caverley uses to establish his point about “civilians
overriding military COIN recommendations” is the extended 1967 controversy
54. Lodge to Johnson, November 7, 1966, FRUS, 1964–1968, Vol. 4, Doc. 294, p. 806. Westmoreland
points out his disagreement with Lodge over military strategy in his memoirs. As he writes, “The
Ambassador was falling back on a long-standing disagreement with my so-called big-unit war
and wanted the bulk of American troops assigned in support of paciªcation.” Westmoreland is
overstating Lodge’s position, but it is obvious that they were on opposite sides of this debate. See
Westmoreland, A Soldier Reports, p. 211.
55. Johnson to Lodge, November 16, 1966, FRUS, 1964–1968, Vol. 4, Doc. 310, p. 848. The president’s reference to the fact that Westmoreland was already making these troop dispositions shows
that he already had the authority to implement the concept he put forward in August.
56. Lodge to Rusk, McNamara and Komer, November 17, 1966, quoted in Mike Gravel, ed., The
Pentagon Papers: The Defense Department History of United States Decisionmaking on Vietnam, Vol. 2
(Boston: Beacon, 1971), p. 606.
Who Lost Vietnam? 115
over Westmoreland and the JCS’s request for as many as 200,000 additional
troops. This is an important moment in the history of the Vietnam War because
the request led to a ªerce debate between the Defense Department and the JCS
over the future of both the ground war and the air war. Summarizing this debate is not easy given the numerous points of view and the duration of the
debate, which lasted from March to July 1967.57 More important, the objectives
that Westmoreland and the JCS hoped to achieve with the additional forces are
hard to discern. When Westmoreland and JCS Chairman Gen. Earle Wheeler
met with Johnson in April 1967 to make the case for an additional two and
one-third divisions, they put forth two distinct arguments. First, Wheeler argued that “[t]hree other matters are bothering the JCS: (a) DRV troop activity
in Cambodia, US troops may be forced to move against these units in Cambodia. (b) DRV troop activity in Laos. US troops may be forced to move against
these units. (c) Possible invasion of North Vietnam. We may wish to take offensive action against the DRV with ground troops.”58 When asked speciªcally by
Johnson what he would do with additional forces, Wheeler responded in a
manner that suggested an interest in bolstering paciªcation in areas such as
the Mekong Delta. MACV’s request for additional forces in the Mekong Delta,
however, did not suggest that U.S. forces would operate differently from those
in other areas of Vietnam: “Primary emphasis will be given to destroying VC
main and guerrilla units and their bases, to intensifying operations to extend
GVN control, to stopping the ºow of food stuffs and material to the enemy
through Cambodia, and to assisting in the ºow of goods to GVN outlets in
Saigon.”59
Caverley makes a useful contribution to the literature by emphasizing that
some civilians in 1967 did argue that additional U.S. forces should not be sent
to Vietnam for purposes of paciªcation. Nevertheless, the troop debate was
not primarily about the philosophical merits of conventional and COIN strategies. Caverley cites John McNaughton’s famous “the philosophy of the war
should be fought out now” document to suggest that it was such a debate, but
it is clear that McNaughton’s actual critique in that document was not about
57. The best sources on the debate sparked by Westmoreland’s March 1967 troop request are
Gravel, The Pentagon Papers, Vol. 4, pp. 424–474, 477–489; and William Conrad Gibbons, The U.S.
Government and the Vietnam War: Executive and Legislative Roles and Relationships, Pt. 4: July 1965–
January 1968 (Princeton, N.J.: Princeton University Press, 1995), pp. 607–669.
58. Notes on Discussion with President Johnson, April 27, 1967, FRUS, 1964–1968, Vol. 5, pp. 349–
352, at p. 351. Col. Charles F. Brower IV makes a persuasive, albeit speculative, argument that
Westmoreland’s troop request was designed to secure support for an invasion of Laos that he had
always thought was necessary. See Brower, “Strategic Reassessment in Vietnam: The
Westmoreland ‘Alternate Strategy’ of 1967–68,” Naval War College Review, Vol. 44, No. 2 (Spring
1991), pp. 20–51.
59. Gravel, The Pentagon Papers, Vol. 4, p. 430.
International Security 35:3 116
paciªcation.60 McNaughton and some other civilian strategists were dubious
about the utility of sizable force increases for any purpose by this point in the
war. The philosophy of the war that McNaughton wanted the administration
to establish was a rejection of the military’s premise that the United States
should do whatever it took to win in Vietnam. If this philosophy was not
clariªed, McNaughton argued that the administration would inevitably be
faced with more troop requests from Westmoreland and Wheeler at a later
date. What McNaughton wanted was for Johnson to set a precise limit on the
number of U.S. ground forces in Vietnam regardless of their purpose. It is this
larger “philosophy of the war” that McNaughton and other civilians wanted
to be fought out with the military in early 1967 and not, as Caverley suggests, a
much narrower philosophical debate over whether the United States should
ªght based on a conventional or a COIN-based paciªcation strategy.61
Even if one accepts this questionable premise that the military was putting
forward a COIN strategy in 1967, it does not prove that Johnson “summarily”
rejected such a strategy. As he often did, and only after an exhaustive process
of deliberation, Johnson took a middle position between the advice he received
from his civilian advisers and military leaders. Opposing the views of some of
his civilian advisers, the president agreed in July 1967 to raise the total number
of U.S. forces in Vietnam from 470,000 to 520,000.62 Although Westmoreland
would undoubtedly have liked to have had all of his requests granted, as any
ªeld commander would, he stated that he was “delighted with the outcome of
the deliberations.”63
Public Opinion and the Johnson Administration
Public opinion is the central component of Caverley’s thesis about the conduct
of the Vietnam War. Caverley argues that the Johnson administration knew
60. McNaughton to McNamara, May 6, 1967, FRUS, 1964–1968, Vol. 5, Doc. 161, pp. 381–383. The
quotations that Caverley attributes to McNaughton on p. 151 of his article are not to be found in
the “philosophy of war” document. The quotations, one of which should be attributed to Wheeler
rather than to McNaughton, can be found in the April 27 discussion with Johnson and
McNamara’s equally important “Draft Presidential Memorandum” of May 19, 1967.
61. In their memoirs, the two primary antagonists in this debate—McNamara and
Westmoreland—did not even mention a dispute over devoting more resources to paciªcation.
Westmoreland emphasized that he wanted the troops for possible actions into Laos, Cambodia,
and North Vietnam, and McNamara emphasized precisely these operations in justifying his opposition. See Robert McNamara, In Retrospect: The Tragedy and Lessons of Vietnam (New York: Random
House, 1995), pp. 264–271; and Westmoreland, A Soldier Reports, pp. 227–228. For a similar view of
what this debate entailed, see Lyndon Johnson, The Vantage Point (New York: Holt, Rinehart and
Winston, 1971), pp. 366–370.
62. Notes of Meeting, July 12, 1967, FRUS, 1964–1968, Vol. 5, Doc. 238, pp. 600–609.
63. Notes of Meeting, July 13, 1967, FRUS, 1964–1968, Vol. 5, Doc. 240, pp. 611–614, at p. 613.
Who Lost Vietnam? 117
that it was pursuing a ºawed strategy in Vietnam but had little or no choice
given the preferences of the average voter for expending treasure rather than
blood. Demonstrating that Johnson and his advisers rejected a more effective
COIN strategy on domestic political grounds, and against the expert advice of
the military, is the crucial thread that connects all of Caverley’s arguments
about the war. It is not by accident that he concludes, “[C]ivilian ofªcials in
Lyndon Johnson’s administration—and ultimately the American public—
played an essential role in the selection of a capital-intensive strategy to ªght
insurgents in the Vietnam War.”64
Assessing the causal impact of public opinion on policymaking over an extended period of time is always difªcult. How a question is worded, when it is
asked, and the various choices offered to the respondent all need to be taken
into account. Linking public opinion to the speciªc decisions of an administration is even more difªcult. Policymakers and presidential advisers can always
attribute a decision reached on the basis of other considerations to the importance of public opinion. For these reasons, reaching judgments about the
inºuence of public opinion should always be a careful process. No one would
deny that Johnson, a consummate politician, was deeply concerned with public opinion. Similarly, no one would deny that Johnson’s advisers were presumably conscious of public opinion and how their advice would affect the
political fortunes of the president.
Nevertheless, an examination of relevant secondary sources and the historical evidence does not support Caverley’s contention that public opinion explains why the Johnson administration pursued particular policies in Vietnam.
As Lawrence Jacobs and Robert Shapiro argue, Johnson’s Vietnam policy “was
based not on public opinion but on his administration’s independent and
expert assessment of America’s strategic interests in Vietnam.” Jacobs
and Shapiro also reject the core argument of Caverley’s thesis, arguing that
“Johnson was unresponsive to public opinion both when he took public positions and when he made decisions on bombing and troop levels.”65 Even
64. Caverley, “The Myth of Military Myopia,” p. 155.
65. Lawrence R. Jacobs and Robert Y. Shapiro, “Lyndon Johnson, Vietnam, and Public Opinion:
Rethinking Realist Theory of Leadership,” Presidential Studies Quarterly, Vol. 29, No. 3 (September
1999), pp. 600, 605. See also Brandon Rottinghaus, “Following the ‘Mail Hawks’: Alternative Measures of Public Opinion on Vietnam in the Johnson White House,” Public Opinion Quarterly, Vol. 71,
No. 3 (Fall 2007), pp. 367–391. Rottinghaus’s analysis of mail received by the White House provides some minor support for Caverley’s argument about responsiveness to hawkish opinion, but
he also agrees with Jacobs and Shapiro that public opinion surveys had little effect. All of
Caverley’s measures of public opinion are the same surveys that scholars in this area have found
to have had little or no impact on Johnson’s decisions.
International Security 35:3 118
sources that Caverley relies on to make minor factual points, such as an extensive 1985 RAND study, explicitly reject the thrust of his argument about administration perceptions of casualties and their impact on policy. As the
authors of the study write, “During the debates in June and July [1965], George
Ball, McGeorge Bundy, William Bundy, McNamara, and Johnson himself
at least mentioned the possibility of increased U.S. casualties resulting in undesirable pressures or problems. Yet these arguments were ultimately discounted, rejected, or not considered sufªciently compelling to seriously affect
policy. There are few indications that the casualty question was a critical consideration inºuencing any major aspect of the decision process in the spring or
summer of 1965.”66
christmas bombing pause
Two examples illustrate Caverley’s tendency to reach conclusions about public
opinion’s impact on decisionmaking that are at variance with readily accessible evidence.
The ªrst example is Caverley’s use of a January 1966 phone call between
Johnson and McNamara that he claims “epitomized” the “perceived strong
constraints from the public” on issues related to the use of force in Vietnam.67
Caverley correctly notes that McNamara expressed concern about how a further extension of the Christmas bombing pause might become a dangerous
campaign issue in the midterm elections of 1966. He does not point out, however, that in the same phone call McNamara and Johnson began with a conversation about how the results of a recent public opinion poll favored the
bombing pause.68 Much to Johnson’s surprise, a Harris public opinion poll
published on January 3, 1966, indicated that 59 percent of the American public
supported a pause in bombing while only 33 percent were opposed; 73 percent
of the public actually favored a temporary cease-ªre.69 When the president expressed his amazement at “the minimum amount of hell-raising that’s come
forth on the pause,” McNamara said that he was not surprised because “the
66. Mark Lorrell and Charles Kelley Jr., with the assistance of Deborah Hensler, Casualties, Public
Opinion, and Presidential Policy during the Vietnam War, No. R-3060-AF (Santa Monica, Calif.:
RAND, March 1985), p. 54. One could hardly imagine a better source to support Caverley’s thesis,
given that the authors interviewed virtually all of the senior civilian ofªcials of the Johnson administration and asked them speciªcally about concerns over casualty rates and their inºuence on
Vietnam policy.
67. Caverley, “The Myth of Military Myopia,” p. 139.
68. Telephone conversation between Johnson and McNamara, January 17, 1966, FRUS, 1964–1968,
Vol. 4, Doc. 26, pp. 74–80.
69. The poll numbers related to the bombing pause can be found in Gibbons, The U.S. Government
and the Vietnam War, p. 128.
Who Lost Vietnam? 119
Republicans read the same poll the rest of us did that showed the wives and
mothers of this country don’t want war.”70 In this case, the perceived constraints of a public hawkish on bombing were not strong enough to prevent either Johnson or McNamara from implementing the bombing pause or for
continuing it for an additional two weeks after the phone call cited by
Caverley. The public’s desire at the end of January 1966 not to continue the
bombing pause was welcome news; indeed the president was counting all
along on greater public support for the war once the pause came to an end. In
fact, Johnson thought that the Christmas bombing pause would make it easier
for him to extract both blood and treasure from the American people. As he
told the National Security Council on January 5, 1966, “We have a better basis
to call on the U.S. people not only for their sons, but also their treasure.
Americans feel better if they know we have gone the last mile even if we have
grave doubts about doing so. The basis for a supplemental budget to pay the
increased costs of the war has been laid.”71
the mcnamara line
The second example of Caverley’s overemphasis of the importance of public
opinion on the strategic choices of the administration is revealed in his discussion of the so-called McNamara Line. For Caverley, McNamara’s commitment
to the construction of a technological anti-inªltration barrier in Vietnam is yet
another example of an inefªcient method of warfare adopted by an administration always looking to substitute capital for labor. He argues that despite
the military’s opposition to the proposal, “McNamara and other politically oriented politicians” promoted it because the barrier was supposedly popular
with the public, and its “development was designed with public opinion in
mind.”72 Rostow, presumably one of those politically oriented ofªcials, allegedly argued that the popularity of the barrier could be used to temper public
discontent with any reduction in bombing: “Surfacing the concept of the barrier may be critical to that turnaround [in public opinion].”73
There is a scholarly literature on the McNamara Line, but Caverley does not
cite it in his article.74 Those interested in the barrier concept may want to ex70. Telephone conversation between Johnson and McNamara, p. 75.
71. Summary Notes of the 555th Meeting of the National Security Council, January 5, 1966, FRUS,
1964–1968, Vol. 4, Doc. 7, pp. 21–22.
72. Caverley, “The Myth of Military Myopia,” pp. 144–145. This explanation raises the question of
why McNamara’s political orientation did not also prevent him from turning against the ostensibly popular bombing policy.
73. Ibid., p. 145.
74. For discussions of the McNamara Line, see Paul Dickson, The Electronic Battleªeld
International Security 35:3 120
amine this scholarship for evidence relating to the role of public opinion in the
development and design of the McNamara Line, but they will ªnd hardly any
mention of it. The most important rationale for the development of the barrier
was that it would enable the administration to reduce the bombing of North
Vietnam, which McNamara and the scientists who proposed it viewed as an
inefªcient method of reducing inªltration. Given this silence, Caverley’s argument must turn on the primary documents he cites. He does cite McNamara’s
approval of the barrier project in September 1966 and his subsequent recommendation to Johnson in October that it should be considered a central element of future strategy in Vietnam, but neither of those crucial documents
contains a word about public opinion.75 Caverley’s best evidence would presumably be the document from Rostow about “surfacing the barrier,” but he
can make his case only by misinterpreting the wording of the document. In
fact, Rostow did not argue that the barrier would be critical to a “turnaround
[in public opinion].” Rather, he stated that surfacing the barrier would be one
component among many in a “turn-around in policy.” Policy and public opinion are surely not the same thing, and by altering the wording of Rostow’s
memo in a manner conducive to his thesis, Caverley is recasting a memo that
is almost exclusively about policy into one focused on public opinion.76
Was the McNamara Line concept of a technological barrier enthusiastically
welcomed by the American public? Caverley cites a single Harris poll in May
1967 that suggests it was more popular than other possible options for “stepping up our military effort,” but I doubt many scholars would view this
(Bloomington: Indiana University Press, 1976); Seymour J. Deitchman, “The ‘Electronic Battleªeld’
in the Vietnam War,” Journal of Military History, Vol. 72, No. 3 (July 2008), pp. 869–887; Christopher
P. Twomey, “The McNamara Line and the Turning Point for Civilian Scientist-Advisers in American Defence Policy, 1966–1968,” Minerva, Vol. 37, No. 3 (September 1999), pp. 235–258; John T.
Correll, “Igloo White,” Air Force Magazine, Vol. 87, No. 11 (November 2004), pp. 56–61; and Peter
Brush, “The Story behind the McNamara Line,” Vietnam (February 1996), pp. 18–24.
75. See McNamara to Lt. General Starbird, September 15, 1966; and McNamara to Johnson, October 14, 1966, both in FRUS, 1964–1968, Vol. 4, Docs. 235, 268, pp. 635–637, 727–735.
76. Rostow to Johnson, May 6, 1967, FRUS, 1964–1968, Vol. 5, Doc. 162, pp. 383–388. Caverley distorts the nature of this document from Rostow in several places in his article. He describes it elsewhere as a memo concerned with “explicitly addressing public responses to escalation strategies”
and as “a memo on ªnding a way to reduce bombing in a manner ‘acceptable to our own people.’”
Caverley, “The Myth of Military Myopia,” pp. 148 and 145, respectively. The document in question, one readily accessible to readers of International Security, is a typical ªve-page Rostow memo
consistent with the analytical and aggressive spirit of virtually all of his memos on the war. His
covering letter to Johnson states, “Herewith thoughts on the alternatives that face us in Viet Nam,”
and that is what he provides to the president. It is only in the next to last paragraph that he raises
even implicitly the idea of making a reduction in bombing “acceptable to our own people”
through a surfacing of the barrier concept and a whole host of other measures. Rostow to Johnson,
pp. 383, 388.
Who Lost Vietnam? 121
ªnding as signiªcant after looking at the variety of other options that respondents were asked to choose from: “(1) Mining the supply harbor in Haiphong
in North Vietnam, even at the risk of sinking Russian ships there, (2) Invading
North Vietnam with troops, (3) Bombing supply lines and air ªelds in China,
and (4) Using atomic ground weapons in the ªghting.”77 It is not surprising
that faced with four grim alternatives, the respondents chose the innocuous
and nonspeciªc option of constructing a “military barrier” in South Vietnam.
Moreover, it is doubtful whether the respondents were offering an opinion on
the highly technological anti-inªltration concept behind the McNamara Line,
which at the time had not been ofªcially revealed to the public. The few published news articles relevant to military barriers from August 1966 to May 1967
did not generally discuss the barrier as a manpower-saving scheme, and it was
sometimes described as a manpower-intensive concept. For example, a
Hanson Baldwin story in the New York Times from April 1967 reported, “As
many as 200,000 to 400,000 men may be required simply to defend this line in
addition to hundreds of thousands of men needed elsewhere to ªght Viet
Cong and North Vietnamese units already established in South Vietnam.”78
One can fairly assume that a single poll question so vague about what it means
by a military barrier, with four grim options as alternatives, cannot qualify as
serious evidence that the American public was enthusiastic about a technological barrier they are assumed to know was capital rather than labor intensive.
Conclusion
Scholarship on the Vietnam War has often been characterized by the search for
blame. Lyndon Johnson, Robert McNamara, William Westmoreland, Walt
Rostow, and others have all spent their time in the historical docket. Despite
having produced indispensible scholarship on the war, both the counterinsurgency and revisionist traditions have spent too much time and energy on prosecuting or exonerating particular civilian or military ofªcials. Caverley’s thesis
continues this regrettable if understandable tendency with a new indictment
against President Johnson, his civilian strategists, and ultimately the American
77. Louis Harris and Associates, 1967, “Harris 1967 Vietnam Survey, No. 1735” (Chapel Hill:
Odum Institute, Louis Harris Data Center, University of North Carolina, 1967), http://hdl.handle
.net/1902.29/H-1735. It is worth noting that the next-most-popular option favored by a majority
of respondents was the presumably labor-intensive option of invading North Vietnam. The
capital-intensive options of bombing China and using atomic ground weapons were not supported by the majority of respondents.
78. Hanson Baldwin, “Vietnam Troop Shifts: Ofªcers Say Moves toward Buffer Zone Do Not End
Need for More Manpower,” New York Times, April 20, 1967.
International Security 35:3 122
people. The purpose of this article is not to defend any of the accused, but to
examine the historical merits of the indictment. Although Johnson and his civilian advisers must bear ultimate responsibility for the Vietnam War, the evidence does not suggest that they instructed or directed the military on how to
ªght a capital-intensive war, and they did not prevent the U.S. military from
adopting a labor-intensive COIN strategy. U.S. military strategy in Vietnam
was a combination of civilian and military decisions, and both groups were responsible for it. Assigning the American people “an essential role” for the selection of the military strategy of the war, as Caverley does, ºies in the face of
common sense and the historical evidence.79
Ultimately, Caverley delivers a theoretical argument and historical synthesis
that does not explain why the United States fought the way it did in Vietnam.
The source of the problem is not hard to discern; his theoretical framework
leads him to simplify and distort the history of the war. There is nothing inherently wrong in approaching the history of the Vietnam War armed with a simple and elegant theory. There is something wrong, however, in making the
complex history of the war conform to that simple and elegant theory. Political
scientists and historians cannot simply locate sentences or phrases in the FRUS
volumes, the Westmoreland papers, or The Pentagon Papers that ªt with their
argument and discard the rest of the readily accessible evidence in the same
document, FRUS volume, or archival box. Historical research cannot be
conducted according to a “search, conªrm, or ignore” strategy. Citing one or
two documents from a lengthy policy discussion in the FRUS volumes does
not absolve any scholar from acknowledging all of the contradictory evidence
that may be contained in other documents in these same volumes. These standards conform to basic understandings of what scholarly research should entail in both history and political science. At a time when scholars are looking to
the Vietnam War as a source for “lessons” relevant to an era of renewed
emphasis on counterinsurgency, it is more crucial than ever that political scientists ground their theories in a solid historical foundation.
A close examination of the history of the Vietnam War suggests another conclusion that would probably be deeply unsatisfying to many historians and
political scientists: the war was unwinnable at an acceptable cost by the summer of 1965, regardless of the type of military strategy adopted by the United
States. The belief that U.S. material power was so great that the country’s defeat must have been the result of human error is understandable, but such a
79. Caverley, “The Myth of Military Myopia,” p. 155.
Who Lost Vietnam? 123
belief overestimates the ability of the United States to control events in places
such as Vietnam. It is highly unlikely that the United States, faced with an experienced and brutal adversary determined to prevail, could have triumphed
with smarter policies or a greater expenditure of blood and treasure. As Col.
Gian Gentile argues, “The United States lost the war in Vietnam. And we lost it
for reasons having less to do with tactics than with the will, perseverance, cohesion, indigenous support, and sheer determination of the other side, coupled with the absence of any of those things on the American side.”80 Gentile’s
perspective offers a way out of the ongoing Vietnam history wars because it
takes scholars’ analytical focus beyond the essential but still narrow focus on
the decisions made by the political and military leadership of the United
States.
Effective counterinsurgency by the United States will always be difªcult because success can be achieved only through a partnership with the local government. As the authors of the U.S. Army/Marine Corps Field Manual
suggest, “U.S. forces committed to a COIN effort are there to assist a HN [host
nation] government. The long-term goal is to leave a government able to stand
by itself. In the end, the host nation has to win on its own. Achieving this requires development of viable local leaders and institutions. U.S. forces and
agencies can help, but HN elements must accept responsibilities to achieve real
victory.”81 This is true in Afghanistan, and it was equally true in Vietnam. The
blood and treasure of the United States is not a viable substitute for the local
forces necessary to combat insurgencies. If there is one crucial lesson to take
away from the history of the Vietnam War, it is to remember Westmoreland’s
principle that the government that the United States is trying to assist in combating an insurgency must ultimately provide security and the prospects of a
better life for its people.
80. Col. Gian P. Gentile, “A (Slightly) Better War: A Narrative and Its Defects,” World Affairs,
Vol. 171, No. 1 (Summer 2008), p. 62.
81. The U.S. Army/Marines Corps Counterinsurgency Field Manual, p. 47.
Explaining U.S. Military Strategy in Vietnam Thinking Clearly about
Causation
Jonathan D. Caverley
International Security, Volume 35, Number 3, Winter 2010/11,
pp. 124-143 (Article)
Published by The MIT Press
For additional information about this article
http://muse.jhu.edu/journals/ins/summary/v035/35.3.caverley.html
Access Provided by University of Arizona at 08/11/11 4:40PM GMT
Explaining U.S. Military Strategy in Vietnam
Explaining U.S. Military
Strategy in Vietnam
Jonathan D. Caverley
Thinking Clearly about Causation
W
hat would cause a
democracy to conduct a ºawed counterinsurgency campaign? What would
lead a democracy to fail in a war against an insurgency? Why would a democracy choose war despite its leaders knowing it will ªght in a manner making
failure more likely? I offered answers to these closely related but crucially distinct questions in “The Myth of Military Myopia: Democracy, Small Wars, and
Vietnam.”1
In “The Myth of Military Myopia,” I argued that leaders of democracies shift
the burden of providing for their nations’ defense onto the rich by substituting
capital (armor, artillery, airpower, etc.) for military labor, thus shielding the relatively less wealthy “median voter” from the costs of defense and of war. Because the costs of ªghting an insurgency with ªrepower are relatively low for
the median voter compared to the more labor-intensive, population security–
oriented approach that is generally recognized to be more effective, rationally
she will favor the former’s use despite the diminished prospects of victory.
In “Myth of Military Myopia,” I employed this deductive argument, which I
call “cost distribution theory,” to examine the case of the Vietnam War and
shed light on an as-yet unresolved historical puzzle: Why did U.S. forces conduct a ªrepower-intensive campaign against the Vietcong (VC) insurgency in
South Vietnam, even as it appeared not to work? I argued that President
Lyndon Johnson and his administration ensured that the U.S. military pursued
a strategy that emphasized the ªght against conventional enemy units and relied on the use of ªrepower for the ªght against insurgents. Furthermore, I
drew on both primary and secondary sources to show that Johnson and his civilian aides were very much aware that although members of the administration considered this strategy ineffective against insurgencies, it was politically
popular in the United States. My argument offers a more compelling account
of the ºawed U.S. counterinsurgency strategy in Vietnam than “military myopia” explanations, which place blame for the United States’ failure on the U.S.
Jonathan Caverley is Assistant Professor of Political Science at Northwestern University.
The author would like to thank Jennifer Light, James Mahoney, Nuno Monteiro, Michael Noonan,
Elizabeth Saunders, Frank Smith, and especially John Schuessler. Lexi Neame provided invaluable
research assistance.
1. Jonathan D. Caverley, “The Myth of Military Myopia: Democracy, Small Wars, and Vietnam,”
International Security, Vol. 34, No. 3 (Winter 2009/10), pp. 119–157.
International Security, Vol. 35, No. 3 (Winter 2010/11), pp. 124–143
© 2010 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
124
Explaining U.S. Military Strategy in Vietnam 125
military’s intrinsic culture, its bureaucratic incentives, or its inability to learn
from experience.2
James McAllister’s response to my article does not challenge the deductive
theory but instead makes three main arguments against the empirical claim
above.3 First, McAllister cites the large amount of primary and secondary
sources that establishes the U.S. military’s disdain for labor-intensive counterinsurgency (COIN), and in that process suggests that my case rests on cherrypicked data.4 Second, he contends that all U.S. decisionmakers, both civilian
and military, understood the centrality of the Government of South Vietnam
(GVN) to a successful war outcome. He uses this evidence to explain why
these decisionmakers preferred to avoid further Americanizing the conºict,
and to suggest that my argument regarding the U.S. military’s poor counterinsurgency efforts fails to capture the roots of the U.S. defeat. Third, McAllister
ªnds unpersuasive the evidence I presented suggesting that, when establishing its counterinsurgency strategy, the Johnson administration considered
voter preferences.
The logic and evidence McAllister uses to support his positions do little
damage to my argument. In this article, I explain why. In the ªrst section, I
show how my argument ªts within the mainstream of Vietnam War historiography, and suggest that deductive reasoning can play an important role in addressing puzzles remaining in this body of knowledge. In the second section, I
show that because McAllister does not distinguish between necessary and
sufªcient causal explanations, his evidence is of little use for testing cost distribution theory against arguments resting on military myopia. In the third section, I address McAllister’s challenge to my evidence that the Johnson
administration expected, encouraged, and even directed its uniformed subordinates to take a capital-intensive approach to counterinsurgency. This section
also examines McAllister’s claim that my linking voter preferences to military
doctrine is unsubstantiated. In the fourth section, I assess McAllister’s conclusion that my efforts to suggest policy implications stemming from my argument are for naught because, in his view, the true cause of the United States’
2. John A. Nagl, Learning to Eat Soup with a Knife: Counterinsurgency Lessons from Malaya and Vietnam (Chicago: University of Chicago Press, 2005), p. 115; Andrew F. Krepinevich Jr., The Army and
Vietnam (Baltimore, Md.: Johns Hopkins University Press, 1986; and Larry E. Cable, Conºict of
Myths: The Development of American Counterinsurgency Doctrine and the Vietnam War (New York:
New York University Press, 1986).
3. James McAllister, “Who Lost Vietnam? Soldiers, Civilians, and U.S. Military Strategy in Vietnam,” International Security, Vol. 35, No. 3 (Winter 2010/11), pp. 95–123.
4. I use “COIN” to distinguish the labor-intensive approach from all other “counterinsurgency”
strategies.
International Security 35:3 126
failure in Vietnam was not a poor strategy but the GVN’s inability to fend for
itself against its enemies. That is, the bulk of McAllister’s article defends the
military myopia thesis, only to conclude that it did not matter for the war’s
outcome. I therefore reiterate the need to distinguish between necessary and
sufªcient causes to explore cost distribution theory’s implications beyond the
study of the Vietnam War. Here again, McAllister’s alternative explanation
and evidence undermine neither my causal logic nor my justiªcation for focusing on counterinsurgency doctrine as a source of failure in small wars.
Theory and the Historical Puzzle of Civilian Noninterference
“The Myth of Military Myopia” challenges an explanation for the U.S. defeat
in Vietnam, not the historical data underpinning that explanation. With the
statement, “If accurate, [Caverley’s] argument would render entire shelves of
books on the Vietnam War obsolete,” McAllister creates a straw man.5 By giving my theory’s historiographic implications too much credit, McAllister gives
my evidence too little. Here, I review how my argument ªts into the literature
on the Vietnam War, and how deductive causal explanations can play a valuable role in the study of history more generally.
Like supporters of the military myopia explanation for U.S. failure in
Vietnam, I concluded that the United States focused on ªghting the enemy’s
main forces rather than insurgents, especially in the pivotal years of 1966 and
1967, a defensible if not undisputed position. My claim that the United States
pursued a counterproductive, ªrepower-intensive counterinsurgency is even
less debatable, as McAllister acknowledges.6 I do not challenge the wellestablished ªnding that Johnson and his civilian advisers directly managed the
ground war much less than they did the air war.
Any contribution to the historical literature I make results from providing an
answer to a question rarely addressed in studies of the Vietnam War: Why was
an administration that was willing to directly manage most other aspects of
the war effort unwilling to closely supervise a ground campaign that appeared
to be failing?7 A deductive theory based on simple assumptions about voters
5. McAllister, “Who Lost Vietnam?” p. 110.
6. Ibid.
7. I am hardly the ªrst to argue that the Johnson administration rejected a more intensive
paciªcation strategy. See, for example, Robert Buzzanco, Masters of War: Military Dissent and Politics in the Vietnam Era (New York: Cambridge University Press, 1996), p. 252. See also Richard A.
Hunt, Paciªcation: The American Struggle for Vietnam’s Hearts and Minds (Boulder, Colo.: Westview,
1995), p. 79; and Elizabeth N. Saunders, Leaders at War: How Presidents Shape Military Interventions
(Ithaca, N.Y.: Cornell University Press, 2011), chap. 5.
Explaining U.S. Military Strategy in Vietnam 127
in democracies and civilian supremacy over the military can suggest causes
that historians have left unaddressed. In the equilibrium predicted by such a
theory, little evidence may exist of civilians directing the military explicitly.8
Inductive approaches will therefore discount the causal link between civilian
leaders’ preferences and the conduct of the military. Agreement by the Johnson
administration and the U.S. military on the United States’ counterinsurgency
strategy, however, is not evidence that the military controlled this strategy. It is
for this reason that considerable weight should be given to my ªnding that,
when civilian leaders and the military disagreed, the former emphasized main
forces and ªrepower more than the latter.
Counterinsurgency, Military Myopia, and Civilian Preferences
McAllister correctly states that, empirically, my “main argument is that
Westmoreland and the military repeatedly tried to convince civilian ofªcials to
adopt a labor-intensive COIN strategy and that ‘the president summarily rejected the COIN option on multiple occasions.’”9 McAllister claims that I am
wrong in asserting that the exclusion of U.S. forces from paciªcation missions
“was promoted solely by civilian policymakers against the better advice or recommendations of the military.”10 This would be a grave error, had I actually
made this argument.
In “The Myth of Military Myopia,” I provided several examples of civilians
in the Johnson administration directing the U.S. military to use ªrepower instead of more ground forces in the war in Vietnam, but at no point did I deny
that Westmoreland largely shared this preference. I do not suggest otherwise
because (1) doing so would be wrong, and (2) this shared preference does
nothing to contradict my theory. McAllister’s emphasis on Westmoreland’s enthusiasm for ªrepower and conventional combat in rebuttal to my argument
belies a misunderstanding of cost distribution theory’s implications. To show
why, I revisit my theory’s causal logic.
Military myopia explanations of the Vietnam War take a largely inductive
approach. These accounts claim to provide a necessary and sufªcient reason
for the ºawed U.S. counterinsurgency effort in Vietnam, one based on characteristics inherent to the U.S. military. By contrast, cost distribution theory is a
8. For the purposes of this article, I use “civilians” to refer to members of the Johnson
administration.
9. McAllister, “Who Lost Vietnam?” p. 109.
10. Caverley, “The Myth of Military Myopia,” p. 146.
International Security 35:3 128
deductive theory about public and government preferences. This theory posits
that, in democracies, civilian leaders’ approval, and ultimately the voters’ support, is necessary for both ºawed counterinsurgency and military myopia.11
Thus, not only are civilian preferences an important mechanism leading to
poor counterinsurgency, but to a large extent, they make spurious the role
played by military myopia. My article denied neither military myopia’s existence nor its effect on warªghting in Vietnam. Instead, it sought to establish
“both military myopia’s limitations and its need to be nested within a theory
of civilian leaders and the public that elects them.”12 In hindsight, a better (if
longer) title for my article might have begun with “The Myth of Military
Myopia’s Causal Power.”
Figure 1 depicts the relationship of cost distribution and military myopia
theories to the dependent variable of ºawed counterinsurgency. Military myopia elaborates on causal relationship 1, where the military’s predilection for
conventional and ªrepower-heavy warfare (MM) is a sufªcient cause of a poor,
capital-intensive counterinsurgency (CC). Cost distribution theory offers
a causally prior independent variable, which for simplicity’s sake I call “civilian preferences” (CP), as a necessary condition for MM, or causal relationship
2. My theory suggests that although MM is not necessary for CC, CP is. Were
the military simply a transmission belt for civilian preferences, democracies
would still engage in CC, or causal relationship 3. The causal role of MM is
therefore diminished.13
Evidence of causal relationship 1 does nothing to undermine cost distribution theory. If the military fought the insurgency exactly as I suggested
civilians preferred, its conduct would have been indistinguishable from the evidence McAllister presents. Cost distribution theory is consistent with Richard
Betts’s claim that “army leaders remained less alienated than those in the other
services because they were less adamant than the navy and air force in their
difference with administration strategy and because the President and the
Ofªce of the Secretary of Defense did not restrict or monitor ground tactics on
anything approaching the scale of which they controlled the air war.”14 I suggest that the two parts of Betts’s explanation are closely related.
11. Ibid., p. 155.
12. Ibid., p. 121.
13. For a formal discussion of this logic, see James Mahoney, Erin Kimball, and Kendra L. Koivu,
“The Logic of Historical Explanation in the Social Sciences,” Comparative Political Studies, Vol. 42,
No. 1 (January 2009), pp. 114–146.
14. Richard K. Betts, Soldiers, Statesmen, and Cold War Crises (New York: Columbia University
Press, 1991), p. 11.
Explaining U.S. Military Strategy in Vietnam 129
Figure 1.
My theory assumes the ultimate dominance of civilians in the determination of security policy, and thus agreement between the administration and its
subordinates is to be expected. It would indeed have been odd if President
Johnson had appointed a commander in Vietnam (MACV) who did not share
his views. H.R. McMaster describes the Vietnam-era military as being at its
Cold War nadir in terms of policy inºuence.15 The Kennedy and Johnson administrations ensured their principal military advisers were handpicked “team
men, not gladiators.”16 Beyond appointments, civilian leaders can shape their
uniformed subordinates’ actions in many indirect ways. The Joint Chiefs of
Staff anticipated the Johnson administration’s preferences in developing and
presenting their campaign plans.17 Statements to the American public also
inºuenced and constrained the military; a 1965 Air Force Policy Letter for
Commanders and a 1968 Air University Review article cite Defense Secretary
Robert McNamara’s televised statement that “what the U.S. sought in South
Vietnam was a limited objective, and it would be accomplished with the lowest possible loss of lives and not necessarily with the lowest expenditure of
money.”18 Given this state of civil-military relations, and the Johnson administration’s deep involvement in every other aspect of the war’s conduct, I suggested applying Occam’s razor to its relative lack of interference in the ground
war.
That many in the administration, including President Johnson, understood
the importance of paciªcation in Vietnam does not damage my case either. Indeed, I noted this understanding in my article.19 Both civilians and the military
15. H.R. McMaster, Dereliction of Duty: Lyndon Johnson, Robert McNamara, the Joint Chiefs of Staff,
and the Lies That Led to Vietnam (New York: HarperCollins, 1997).
16. George C. Herring, LBJ and Vietnam: A Different Kind of War (Austin: University of Texas Press,
1994), pp. 29–30.
17. McMaster, Dereliction of Duty, pp. 329–334.
18. Robert McNamara interview, February 8, 1965, quoted in “Air Force Policy Letter for Commanders,” February 15, 1965. References to both the interview and the policy letter are in Robert
M. Kipp, “Counterinsurgency from 30,000 Feet: The B-52 in Vietnam,” Air University Review, Vol.
19, No. 2 (January–February 1968), pp. 10–18.
19. Caverley, “The Myth of Military Myopia,” p. 146.
International Security 35:3 130
recognized the labor-intensive principles of COIN and that ªrepower was a
poor substitute. Nevertheless, they pursued ºawed counterinsurgency. For example, the famous February 1966 Honolulu conference convened by Johnson,
which “clearly endorsed an improved paciªcation effort,” supports my claims
because the concrete instructions given to Westmoreland in Honolulu made attrition “the primary operational objective.”20
To show that civilian preferences (CP) is a superior explanation to military
myopia (MM) requires focusing on relationships 2 and 3 in ªgure 1. The evidence for these relationships may not be plentiful given that cost distribution
theory predicts civil-military harmony in equilibrium. On the other hand, any
ªnding of instances where civilians helped direct, shape, and reinforce the military’s capital-intensive campaign (i.e., relationship 2) means that “cost distribution theory explains more aspects of the war” than does military myopia.21
Despite the existence of the military’s enthusiasm for ªrepower in Vietnam,
linking civilian preferences to counterinsurgency strategy absent a role for
military myopia (relationship 3) is not entirely counterfactual, given the
Johnson administration’s dealings with more COIN-oriented civilians such as
Ambassador Henry Cabot Lodge (who axiomatically did not suffer from military myopia).
Conversely, the most direct way for McAllister to overturn my argument is
to provide evidence that the Johnson administration preferred paciªcation
more strongly than the military did, and yet the military’s strategy of main
force focus, ªrepower, and attrition prevailed. Instead, he provides much
material in support of relationship 1 and challenges some of my evidence for
relationship 2.22
The Civilian Role in the Historical Record
Space precludes addressing all of McAllister’s empirical objections, but given
that my argument neither claims nor requires constant civil-military disagreement, this is unnecessary. I therefore tackle the aspects of McAllister’s rebuttal
that truly bear on my theory (causal relationship 2 in ªgure 1), brieºy highlight evidence that McAllister disregards, and acknowledge aspects where he
20. Michael A. Hennessy, Strategy in Vietnam: The Marines and Revolutionary Warfare in I Corps,
1965–1972 (Westport, Conn.: Praeger, 1997), pp. 82–83.
21. Caverley, “The Myth of Military Myopia,” p. 131.
22. A competing explanation might assign causal priority to MM, but make CP necessary as well.
Deborah D. Avant argues that the “electoral circumstances” of civilians prevented them from inducing the U.S. Army to change its conventional bias. I assign civilians a much more positive role.
Avant, Political Institutions and Military Change: Lessons from Peripheral Wars (Ithaca, N.Y.: Cornell
University Press, 1994).
Explaining U.S. Military Strategy in Vietnam 131
adds corrections and nuance. In so doing, I address his claim that I engaged in
poor scholarship.
“The Myth of Military Myopia” sought to establish three ªndings. First, the
public in a democracy has a preference for spending treasure over blood. Second, the government considers this preference in developing and approving a
capital-intensive strategy. Third, the government must explicitly direct the military to employ this strategy.23 McAllister focuses primarily on this last proposition, makes some observations on the second, and says little about the ªrst. I
will do likewise.
the administration’s inºuence in the ground campaign
McAllister’s case against my article rests largely on the role of the Johnson administration in determining the ground strategy in Vietnam. Contrary to
McAllister’s assertion, my article presented more than “three examples to argue that the administration rejected a labor-intensive COIN approach.”24 For
example, he does not acknowledge Westmoreland’s 1964 recommendation to
maintain an advisory approach in Vietnam rather than initiate bombing or
ground operations, justiªed by Westmoreland’s deep involvement in a “laboratory experiment in paciªcation.”25 Nonetheless, this section concentrates on
the three points addressed by McAllister.
the phase ii debate, 1965. McAllister objects to my interpretation of
National Security Adviser McGeorge Bundy’s September 23, 1965, memo as a
rejection of paciªcation strategy.26 This memo was part of the Johnson administration’s debate over “Phase II” troop deployments and expansion of the
strategic bombing campaign, Operation Rolling Thunder. In my article, I argued that during this debate Bundy and other administration members suggested that if U.S. troops were not going to be used against the enemy’s main
force, they need not be deployed at all.
McAllister correctly notes that Bundy’s memo did not explicitly repudiate
paciªcation.27 Rather, Bundy summarized Westmoreland’s ground war plan,
which prominently features paciªcation, and observed that “the problem
arises as to how we use our substantial ground and air strength effectively
against small-scale harassment-type action, whether we should engage in
paciªcation as opposed to patrolling actively, and whether, indeed, we should
taper off our ground force build-up.” The memo reports the administration’s
23.
24.
25.
26.
27.
Caverley, “The Myth of Military Myopia,” p. 131.
McAllister, “Who Lost Vietnam?” p. 110.
Caverley, “The Myth of Military Myopia,” p. 149. See also McMaster, Dereliction of Duty, p. 188.
McAllister, “Who Lost Vietnam?” p. 110.
Ibid.
International Security 35:3 132
tasking Lodge with developing a plan involving “the concentration of GVN
forces on paciªcation and the reliance on U.S. forces to handle large-scale VC
actions.”28 This notion that more U.S. troops should be sent only to engage in
main force battles also appears in Bundy’s earlier memo recommending that
“we should explicitly and plainly reserve decision about further major deployments. After all, we have not yet had even a company-level engagement with
Viet Cong forces which choose to stand their ground and ªght.”29
Secretary of State Dean Rusk reiterated this plan in a telegram to Lodge:
We need to consider just how we propose to use our greatly increased ground
and air strength, especially the degree to which it can and should be employed
in any wider countryside efforts beyond necessarily slow securing efforts close
to our base areas. (We also question whether and how we can move from patrolling to real paciªcation in these areas—can ARVN [the Army of South
Vietnam] and GVN police take advantage of our nearby strength for this purpose in these areas?) There is even a residual question whether further increases in strength at presently planned pace are wise, or whether we should
in some small degree defer further increases.30
These linked concerns over manpower and paciªcation also appear in
Bundy’s October 26, 1965, draft telegram to Westmoreland and Lodge, which
contained a number of observations prompted by a military brieªng on Phase
II, the ªrst being that “these plans focus sharply upon a dominant ªghting role
for U.S. ground forces. They appear to imply that aggressive operations will be
conducted almost exclusively by U.S. forces.”31
Notes from a July 22, 1965, presidential meeting (one of the few to include
the Joint Chiefs of Staff) provide an additional example of administration resistance to paciªcation. They record Marine Commandant Wallace Greene arguing, “The enclave concept will work. Would like to introduce enough Marines
to do this.” McNamara observes that Greene is asking for “men over and
above the Westmoreland request.” Johnson responds, “Then you will need
80,000 more Marines to carry this out.”32 More Marines for paciªcation were
not forthcoming.
28. Foreign Relations of the United States (FRUS), 1964–1968, Johnson Administration, Vol. 3, Doc. 151,
http://www.state.gov/www/about_state/history/vol_iii/index.html. McNamara’s Phase II descriptions barely mention paciªcation. See ibid., Doc. 67, “Use of Forces” section; and Doc. 149. Because McAllister frequently criticizes my interpretation of primary sources, this article will cite
material that can be easily found and interpreted by interested readers.
29. Ibid., Doc. 83.
30. Ibid., Doc. 141.
31. Ibid., Doc. 183. Westmoreland complains of these interfering cables in William C.
Westmoreland, A Soldier Reports (New York: Da Capo, 1989), p. 161.
32. FRUS, 1964–1968, Vol. 3, Doc. 76. The enclaves concept called for concentrating U.S. forces in
Explaining U.S. Military Strategy in Vietnam 133
Westmoreland may have largely agreed with the civilian leadership’s focus
on enemy main forces, but the civilians were well informed on the strategic
options, clearly in charge, and concerned whether Westmoreland and others
understood their preferred strategy. The memos discussed above support civilian preferences over military myopia (causal relationship 2 in ªgure 1); they
also support causal relationship 3, because they show that the administration
pushed the less ªrepower-centric Lodge in a similar direction.
westmoreland’s conops, 1966. McAllister disagrees that Westmoreland’s
description of his 1966 campaign plan to his civilian superiors represented
a shift toward paciªcation. Once again, his criticism centers on showing
evidence of military myopia (causal relationship 1), whereas the plan’s signiªcance for theory testing is in civilian leaders’ negative reaction to
Westmoreland’s explicit intention for “a signiªcant number of the U.S./
Free World Maneuver Battalions” to be involved in paciªcation (relationship 2).33 Subsequent decisions and directives by the administration favoring
Westmoreland’s plan over Lodge’s much more paciªcation-oriented plan further underscore that the military was not the ultimate cause of the ªrepowerintensive campaign of 1966.
National Security Adviser Walt Rostow noted the Westmoreland plan’s shift
“towards paciªcation,” observing that it “underlines the need to mount a maximum political campaign, overt and covert, designed to defect VC and start
Saigon VC negotiations . . . required to match Westmoreland’s military plan
which is, clearly, in the right direction; although he and Lodge must engage
[Prime Minister] Ky and the ARVN fully if it is to work.” Johnson’s order to
have his principal paciªcation adviser, Robert Komer, “spark this inspiration”
(and Komer’s subsequent focus on the GVN throughout 1966) referred to getting the GVN rather than the U.S. military more heavily involved in paciªcation.34 On the other hand, the president’s directive that Westmoreland should
not consider his concept of operations approved was clearly prompted by retired general and former Ambassador to Vietnam Maxwell Taylor’s scathing
analysis of the domestic political problems inherent in Westmoreland’s shifting emphasis, however modest, toward paciªcation.35
Whereas Taylor considered Westmoreland’s paciªcation plans overly aggressive, Lodge believed that they did not go far enough and recommended a
fortiªed population centers and increasing control of the surrounding areas gradually, a strategy
closer to COIN than Westmoreland’s or Johnson’s. Hennessy, Strategy in Vietnam, pp. 74–77.
33. FRUS, 1964–1968, Vol. 4, Doc. 220.
34. Ibid.
35. Ibid., Doc. 221; and Caverley, “The Myth of Military Myopia,” p. 151.
International Security 35:3 134
more ambitious U.S. effort while lambasting Westmoreland’s approach. In
Lodge’s words,
MACV speciªcally states that what it calls “offensive operations” are
conducted so as to create the opportunity to destroy terrorism, that is
“paciªcation”.
But the phrase “offensive operations” is deªned as meaning to “seek out
and destroy”. . . . I believe that the Vietnamese war will certainly never be won
in this way; that the phrase “offensive operations” should be deªned as “split
up the Viet Cong and keep him off balance”; and that U.S. participation in
paciªcation operations should be stepped up.36
Contrary to McAllister’s assertion, this strategic disagreement was far from a
“minor element,” nor is Lodge’s memo “devoted entirely” to a “management
issue” over whether the U.S. embassy in Saigon or MACV would lead the
paciªcation effort.37 According to the FRUS editor, Lodge sent “several similar
communications” including a November 6 telegram wherein “he stated that
the crux of the problem was security, not defective organization, and that the
ªrst priority was more U.S. troops allotted to paciªcation.”38
Johnson favored Westmoreland’s greater preference for ªrepower and emphasis on main force battles over Lodge’s more labor-intensive option. With
characteristic disingenuousness, in the memo below, Johnson threatens to shift
paciªcation responsibilities from Lodge to MACV even as he appears to agree
with the ambassador:
There does not seem to me to be any major difference between your ideas
of what is needed to make paciªcation work, and those of my chief advisers
and myself. Bob McNamara and the Joint Chiefs realize, as does General
Westmoreland on the basis of the dispositions he is increasingly making, that a
limited number of U.S. combat forces must be detailed to be the catalysts for
the Vietnamese.
What worries them is rather that if the U.S. takes over too much of the job,
the ARVN will tend to sit back and let us ªght that “war” too. I’m sure that
you are no more eager than we are to let this happen. As a matter of fact, getting the U.S. military more heavily engaged in refocusing ARVN on the heart
of the matter is one reason why we here have seriously considered charging
MACV with paciªcation. I hope you will ponder whether this is not in the end
the best way to achieve the aim you seek.39
36.
37.
38.
39.
ply
FRUS, 1964–1968, Vol. 4, Doc. 294.
McAllister, “Who Lost Vietnam?” p. 114.
FRUS, 1964–1968, Vol. 4, Doc. 294; and ibid., Doc. 290 n. 2.
Ibid., Doc. 310. See also Hunt, Paciªcation, p. 79. Johnson also understood that the military simhad more resources at its disposal for paciªcation but makes no reference to this here.
Explaining U.S. Military Strategy in Vietnam 135
In this context, McAllister’s selection from this document of Johnson’s observation that, “[i]f showing ARVN the way on paciªcation can take up to ten
percent of our troops, it also deserves the full time attention of some of our
best generals,”40 represents a presidential threat that continued insistence on
this percentage of U.S. troops for paciªcation would result in the military being given the assignment.
philosophy of the war, 1967. McAllister objects to my characterization of
the 1967 policy debate as being over paciªcation when it largely dealt with the
need to head off additional manpower requests from the military.41 The whole
point of my article, however, is that these two issues cannot be separated; civilians continued to favor main force engagements and were prepared to shift
forces away from the “less essential” (McNamara’s phrase) paciªcation mission in an effort to limit new personnel deployments.42 While the main policy
decision was ostensibly over whether the president should approve the Joint
Chiefs’ request for roughly 200,000 more troops or authorize a much smaller
number, Assistant Secretary of Defense John McNaughton’s May 6, 1967,
memo to McNamara argues that limiting troop numbers is not enough because “the strategy falls into the trap that has ensnared us for the past three
years. It actually gives the troops [to MACV] while only praying for their
proper use.”43
To show the Johnson administration’s position on the troops’ “proper use,”
I turn to the person likeliest to recommend the most aggressive COIN
approach the president would tolerate. Robert Komer wrote a memo on
April 24, 1967, “deliberately designed to plead an alternative case.”44 In it, he
directly addresses the use of U.S. forces in the ªrst paragraph of his list of
recommendations,
MACV’s justiªcation for these added forces needs further review. . . . If enemy
main force strength is now levelling off because of high kill ratios, etc., would
the added US forces be used for paciªcation? General [William E.] DePuy
estimates that 50% of US/ROK [Republic of Korea] maneuver battalions are
40. McAllister, “Who Lost Vietnam?” p. 114.
41. Ibid., p. 22.
42. FRUS, 1964–1968, Vol. 5, Doc. 177.
43. FRUS, 1964–1968, Vol. 5, Doc. 161. McAllister rightly notes my incorrect attribution to this May
6 memo of several quotations from McNaughton’s version of a Draft Presidential Memorandum
(DPM). Both documents are described at length in Mike Gravel, ed., The Pentagon Papers: The Defense Department History of United States Decisionmaking on Vietnam, Vol. 4 (Boston: Beacon, 1971),
pp. 477–489. The FRUS volume includes a DPM version sent to the president by McNamara.
FRUS, 1964–1968, Vol. 5, Doc. 177. I thank McAllister sincerely and take full responsibility for this
error.
44. Ibid., Doc. 147.
International Security 35:3 136
already supporting RD [Revolutionary Development, i.e., paciªcation] by
dealing with the “middle war”, the VC main force provincial battalions. How
good are US forces at paciªcation-related tasks, as compared to RVNAF [Republic of Vietnam Armed Forces]? What are the trade-offs? A major US force
commitment to paciªcation also basically changes the nature of our presence
in Vietnam and might force us to stay indeªnitely in strength.45
To “reduce or obviate the need for a major US force increase,” Komer ªrst
recommended “an all-out effort to get more for our money out of” South
Vietnam’s military, including an increase in the number of U.S. advisers by a
mere 1,200, or “the strength of one USMC [U.S. Marine Corps] maneuver battalion.” The other steps involved a minor expansion of civilian paciªcation
personnel, bigger local militias, better intelligence collection, land reform, refugee management, and a transition to an “effective, popularly-based GVN.”
Komer concludes, “The above package could be combined with other US unilateral measures—let’s say a minor force increase to 500,000 [from about
470,000], accelerated emphasis on a barrier, and some increased bombing—to
further optimize its prospects.”46
Johnson and Komer chose Westmoreland’s version of counterinsurgency
over Lodge’s more labor-intensive paciªcation approach, but Westmoreland
also frequently appeared more interested in paciªcation than his civilian masters. Because McAllister relies heavily on Westmoreland’s memoir, I give it the
last word on this subject: “In reality, despite my policy of using American units
to oppose the enemy’s main forces, more American troops were usually engaged on a day-by-day basis, helping weed out local opposition and supporting the paciªcation process, than were engaged in the big ªghts.”47 This
description of Westmoreland’s ground war little resembles the civilian desires
and expectations depicted in the above memos.
the administration’s inºuence on air and armor decisons
Although McAllister focuses primarily on the Johnson administration’s involvement (or lack thereof) in conducting the ground war, he also challenges
my evidence that it played an important role in ensuring that airpower was
used to the maximum extent feasible in South Vietnam, and that civilian policy
decisions shaped the use of armor in Vietnam by the U.S. Army.48 Below, I
brieºy respond to these objections.
45.
46.
47.
48.
Ibid.
Ibid.
Westmoreland, A Soldier Reports, p. 146.
McAllister, “Who Lost Vietnam?” pp. 104–105.
Explaining U.S. Military Strategy in Vietnam 137
The Johnson administration’s involvement in airpower targeting in South
Vietnam is not debatable. A 1966 U.S. Air Force analysis describes the B-52
missions in the South as a “major administration decision.”49 McNamara
clearly instructed the reluctant Joint Chiefs that the South had priority over the
North when determining how to use airpower assets.50 The defense secretary
later mooted that a Rolling Thunder pause would be more acceptable to the
American public if these planes’ missions were shifted to the South.51
McNamara set the number of sorties in South Vietnam (1.2 per aircraft per
day), and B-52 strikes in the South had to be approved up through early 1966
by McNamara, Rusk, and the president.52 Even after the civilian micromanaging subsided, it is unclear what further measures McNamara needed beyond
setting the aircraft and sortie numbers in country, allowing unrestricted air operations in the South, and strongly restricting them in the North to ensure
airpower’s copious use in the South. Westmoreland’s back-channel complaints
to other generals only provide additional support for my case.53
Likewise, contrary to McAllister’s claim, civilian decisions shaped the army
leadership’s use of armor. The analysis by the commanding general of the U.S.
Army’s Armor School states explicitly why it was hardly used in Vietnam
through 1966: “Because of the troop ceiling . . . the severely limited logistical
base, and the many misconceptions about the country, armored units were not
seriously considered for early employment in Vietnam.”54 MACV did not control the number of soldiers at its disposal but could decide which units to deploy. Westmoreland and other army leaders favored less heavily-armored
units because they understood that Vietnam was not a conventional war, and
thought that lighter battalions would be most effective given manpower constraints. This emphasis by the military commanders is hard to square with military myopia’s claims of an obsession with ªrepower rather than soldiers.
The ªrst signiªcant heavy armored unit arrived in Vietnam only in September 1966 after being largely stripped of its main battle tanks.55 According to
49. Jacob Van Staaveren, “USAF Plans and Operations in Southeast Asia, 1965” (Washington,
D.C.: USAF Historical Division Liaison Ofªce, 1966), pp. 39–40.
50. Ibid. See also FRUS, 1964–1968, Vol. 3, Doc. 183. McAllister’s claim that “civilians attempted to
instruct the military to recognize the merits of gradualism and limitations” conºates operations in
the South and North. McAllister, “Who Lost Vietnam?” p. 107. Civilians advocated gradualism
only in the North, despite bombing’s public popularity, to avoid antagonizing China and the
Soviet Union.
51. FRUS, 1964–1968, Vol. 3, Doc. 231.
52. Van Staaveren, “USAF Plans and Operations in Southeast Asia,” p. 46; and William W.
Momyer, Airpower in Three Wars (Maxwell Air Force Base, Ala.: Air University Press, 2003), p. 114.
53. Caverley, “The Myth of Military Myopia,” p. 142.
54. Donn A. Starry, Armored Combat in Vietnam (New York: Arno, 1980), p. 55.
55. Westmoreland’s memoir claims “enthusiasm” for the army’s primary main battle tank, the
M-48 “Patton,” but the facts reported in Starry speak otherwise. This is one of many empirical rea-
International Security 35:3 138
Donn Starry, the impact of the 1967 pro-armor study cited by McAllister “was
somewhat less than many hoped for,” in part because “in November 1966,
Defense Secretary McNamara imposed an absolute troop ceiling on U.S. forces
in Vietnam. This arbitrary ceiling was well below the total number already in
[MACV’s] proposed troop program.” Starry continues, “If more armored
forces were wanted, other units had to be given up in order to get them,” and
MACV apparently did not want to give up any existing units.56 In Starry’s account, armored units became an important factor in battles against the large
1968 enemy offensives and during the subsequent drawdown “because they
provided mobility and ªrepower at far less cost in manpower than any other
type of unit.”57 In sum, civilian manpower decisions drove the army’s shift
from favoring soldiers to deploying larger numbers of tanks.
public preferences and administration decisionmaking
McAllister challenges my linking of public opinion to government preferences
for substituting ªrepower for labor.58 Establishing direct links between polls
and presidential action is a well-recognized problem, and I did not pretend to
show conclusive evidence.59 Instead, in “The Myth of Military Myopia,” I argued that members of the administration believed that ªrepower, and especially airpower, was the most politically acceptable means of prosecuting the
war. While bombing was clearly an inefªcient way to ªght the insurgency,
the American public strongly believed that it was an important replacement
for manpower and a means of protecting soldiers’ lives, and civilian leaders
considered their strategy options accordingly. Here, as before, I focus on the
two examples from my article that McAllister uses to support his claim—a
presidential meeting on January 5, 1966, and a phone call between McNamara
and Johnson on January 17, 1966, both of which addressed the “bombing
pause” (i.e., a temporary cessation of the Rolling Thunder bombing campaign
in North Vietnam)—rather than all the other pieces of primary evidence I marshaled to support my claims.
McAllister uses the January 5 meeting notes to “illustrate Caverley’s ten-
sons why I favored the primary record over Westmoreland’s post facto recollections, a decision
that McAllister ªnds “troubling.” Westmoreland, A Soldier Reports, p. 178; Starry, Armored Combat
in Vietnam, pp. 72–73; and McAllister, “Who Lost Vietnam?” p. 105.
56. Starry, Armored Combat in Vietnam, pp. 86–89.
57. Ibid., p. 137.
58. McAllister, “Who Lost Vietnam?” p. 97.
59. The research McAllister cites is far from conclusive. For example, Lawrence R. Jacobs and Robert Y. Shapiro largely focus on Johnson’s inability to lead public opinion. Jacobs and Shapiro,
“Lyndon Johnson, Vietnam, and Public Opinion: Rethinking Realist Theory of Leadership,” Presidential Studies Quarterly, Vol. 29, No. 3 (September 1999), pp. 592–616.
Explaining U.S. Military Strategy in Vietnam 139
dency to reach conclusions about public opinion’s impact on decisionmaking
that are at variance with readily accessible evidence.”60 Ironically, McAllister
chooses a document almost entirely devoted to Johnson’s ostentatious, global
diplomatic effort known as the “peace offensive,” which emphasized the
bombing pause’s popularity abroad, not in the United States. The notes report
that, following a lengthy brieªng on outreach attempts to foreign leaders, Secretary of State Rusk distinguished explicitly between domestic and international support for the pause: “Our position will erode here if we wait much
longer to resume the bombing but abroad we will lose support if we resume.”
Johnson responds, “The diplomatic offensive boils down to saying that we are
ready to reason this out. One poll shows that 73 percent of the American people wanted us to increase our diplomatic efforts. In the last twelve months, 200
conferences have been held by Secretary Rusk in an attempt to get negotiations
going.”61 The president thus identiªes popular support in the United States
with diplomacy, of which the pause was a component. The connection of the
pause to diplomacy is further established in the January 17 phone call. After
expressing surprise at the pause’s domestic popularity, Johnson asked
McNamara, “[D]o you think we’re going to have a sentiment that will support
our resumption if everybody feels this way about it?” McNamara replies, “I
think so, Mr. President, particularly among the great majority of the people in
this country. I think you’ll ªnd some foreign leaders will criticize you if you resume bombing. . . . But I think the great majority of the people in the country
will believe that you gave them [the North Vietnamese government] a reasonable time, over a month, and there was no movement at all on their part.”62
Johnson’s belief that the American public preferred to achieve its aims in
Vietnam with minimal warªghting is neither surprising nor damaging to my
case. Johnson also believed, however, that he would soon face strong political
pressure to ªght for those aims should diplomacy fail, and that airpower was
the public’s preferred means of doing so. Not coincidentally, December 7, 1965,
notes on Johnson’s deliberation over the pause contain some of his most famous musings on domestic politics and the war: “I think we’ll be spending
more time defending ourselves from hawks than from doves. . . . We’re spending too much time with crybabies. Average fella doesn’t have much respect.
Afraid we’ll lose our own ªghting men. . . . We’ve got a new election here. This
is a priority problem. It comes ahead of poverty & education. It’s a new ball
60. McAllister, “Who Lost Vietnam?” p. 118.
61. FRUS, 1964–1968, Vol. 4, Doc. 7.
62. Ibid., Doc. 26.
International Security 35:3 140
game. 1966 election.”63 As additional evidence, consider Rostow’s September
1966 memo declaring, “Bombing in the North is our equivalent of Viet Cong
guerrilla operations in the South.” The memo’s ªnal paragraph reads in full, “I
add an amateur political judgment: a ‘pause’ during the campaign, without
solid evidence that a move towards peace will promptly follow, could be quite
dangerous during the campaign, as well as providing evidence of over-anxiety
and lack of perseverance to Hanoi.” McNamara’s handwritten note next to this
paragraph reads, “I am inclined to agree that a ‘pause’ prior to November
would be unwise.”64 No plausible interpretation of this paragraph exists other
than that Johnson’s two most important national security aides considered
Rolling Thunder a source of electoral strength in the upcoming midterm
elections.
I conclude this section by returning to another Rostow memo that encapsulates most of my empirical claims about the driving factors of U.S. strategy in
Vietnam but also prompts McAllister’s assertion that by “altering the wording
of Rostow’s memo in a manner conducive to his thesis, Caverley is recasting a
memo that is almost exclusively about policy into one focused on public opinion.”65 McAllister objects to my interpretation of the word “turn-around” in
this memo, and I regret any confusion this caused. Still, my use of this word
does nothing to change the thrust of Rostow’s concluding paragraph:
The turn-around in policy can be managed, over a period of some weeks, in
the context of Buddha’s birthday, etc., fairly easily; but if we get no diplomatic
response in that period—and I do not expect one—and if we set aside option A
(closing the top of the funnel), we shall have to devise a way of presenting our
total policy in Viet Nam in a manner which is consistent with diminished attacks in the Hanoi-Haiphong area; which is honest; and which is acceptable to
our own people. Surfacing the concept of the barrier may be critical to that
turn-around, as will be other measures to tighten inªltration, an improved
ARVN effort in paciªcation, and the provision of additional allied forces to
permit Westy to get on with our limited but real role in paciªcation—notably,
with the defense of I Corps and the hounding of provincial main force units.66
Rostow is clearly discussing the likely public reception of a policy turnaround.
Unlike McAllister, I do not see how in the context of “presenting our total policy” in a manner “acceptable to our own people,” the phrase “surfacing the
63. FRUS, 1964–1968, Vol. 3, Doc. 223. See also ibid., Doc. 215.
64. FRUS, 1964–1968, Vol. 4, Doc. 232.
65. McAllister, “Who Lost Vietnam?” p. 120.
66. FRUS, 1964–1968, Vol. 5, Doc. 162. “Closing the funnel” refers to interdicting the supply lines
to enemy forces in South Vietnam from the North.
Explaining U.S. Military Strategy in Vietnam 141
concept of the barrier” describes the actual use of the anti-inªltration barrier
rather than its revelation to the public.67 McAllister points out that this is a
“typical,” aggressive memo from the president’s national security adviser that
is largely devoted to examining alternative policies in Vietnam, a memo that
concludes with a paragraph addressing public support for the war.68 I agree.
Necessary versus Sufªcient Causes of Failed Counterinsurgencies
In this section I return to the question: Why do democracies fail against insurgencies? McAllister disagrees with my claim that part of the explanation in the
case of Vietnam can be found in the United States’ reliance on a ªrepowerintensive counterinsurgency; indeed he suggests that my article presents an
unfortunate distraction from the true cause of failure, which he argues lies in
GVN incompetence. Rebutting this charge necessitates another discussion of
necessary and sufªcient causation. In his conclusion, McAllister creates another straw man by transforming my argument into a claim that populationcentric COIN is sufªcient for a successful outcome against an insurgency. In
“The Myth of Military Myopia,” I made no such argument.
McAllister, on the other hand, does. He writes, “If there is one crucial lesson to take away from the history of the Vietnam War, it is to remember
Westmoreland’s principle that the government that the United States is trying
to assist in combating an insurgency must ultimately provide security and the
prospects of a better life for its people.”69 To support this assertion, McAllister
employs evidence that both the Johnson administration and the U.S. military
feared the Americanization of the war, which they believed would take pressure off the South Vietnamese forces to reform and aggressively combat the insurgency. These American decisionmakers understood that the insurgents
would ultimately win if a competent South Vietnamese government did not
emerge.70
McAllister’s argument suffers from two ºaws in causal logic. First, one
cannot consider the weakness of South Vietnam’s government and the coun67. McAllister, “Who Lost Vietnam?” pp. 119, 120 n. 76. The editor of The Pentagon Papers shared
my assessment. Gravel, The Pentagon Papers, Vol. 4, p. 477. Regarding the barrier, Paul N. Edwards
describes the McNamara Line as “a microcosmic version of the whole United States approach to
the Vietnam War” in which the operations of the entire U.S. military were centralized under civilian control. Edwards, The Closed World: Computers and the Politics of Discourse in Cold War America
(Cambridge, Mass.: MIT Press, 1996), p. 5.
68. McAllister, “Who Lost Vietnam?” p. 120 n. 76.
69. Ibid., p. 123.
70. Ibid., p. 108.
International Security 35:3 142
terinsurgency efforts of the U.S. military as separate factors. Tragically, while
third-party counterinsurgency efforts can lead to increased incumbent government incompetence, the former is also endogenous to the latter; a sufªciently
capable government would obviate the need for much counterinsurgency.
Second, the realization that multiple routes to failure in Vietnam existed
undermines neither the power of my argument nor the justiªcation for my focusing on a single cause. Both the straw man argument that poor COIN is necessary for failure and McAllister’s own “one crucial lesson” are empirically
false. Successful counterinsurgency has many requirements, including establishing government competence, instigating paciªcation, checking the enemy
main force, and curtailing state sponsors of the insurgency. Each of these antecedent variables is necessary but not sufªcient for a victory against insurgents.
Conversely, while failure against an insurgency cannot be an orphan, it needs
only one parent. As Leslie Gelb and Richard Betts put it, “The United States
could not win by paciªcation alone, but it could not win without paciªcation.”71 Moreover, I argue that ºawed counterinsurgency is the one sufªcient
condition for failure to which democracies, such as the United States, may be
uniquely prone.
Conclusion
In “The Myth of Military Myopia,” I sought to apply to the Vietnam War a theory of cost distribution within democracies, where military capitalization
serves as a means of redistributing the costs of conºict away from the median
voter. I did this because investigations of the decision “to lose Vietnam slowly”
still contain large explanatory gaps.72 I ªlled one of these gaps by arguing that,
acting on a perception of the American public’s preferences, the Johnson administration deliberately, knowingly, and rationally chose a capital-intensive
strategy that made failure against the Vietcong more likely. McAllister’s charge
that I ascribed the U.S. defeat in Vietnam to “human error” belies a misunderstanding of cost distribution theory’s implications and my article’s aims.73
Although McAllister ªnds me insufªciently like a historian in my collection
and presentation of evidence, his conclusion ªnds me too much like one because of what he claims is my “evident hostility” and “indictment against
71. Leslie H. Gelb and Richard K. Betts, The Irony of Vietnam: The System Worked (Washington, D.C.:
Brookings Institution Press, 1979), p. 251.
72. Larry Berman, Planning a Tragedy: The Americanization of the War in Vietnam (New York: W.W.
Norton, 1982), p. 124.
73. McAllister, “Who Lost Vietnam?” p. 122.
Explaining U.S. Military Strategy in Vietnam 143
President Johnson, his civilian strategists, and ultimately the American people” for the U.S. defeat in Vietnam.74 McAllister offers an alternative to
this pathological practice in which I apparently engage, “a way out of the ongoing Vietnam history wars because it takes scholars’ analytical focus beyond
the essential but still narrow focus on the decisions made by the political and
military leadership of the United States.”75 McAllister’s formulation of this
“way out”—the United States lost Vietnam because it was “unwinnable at an
acceptable cost”—misses my point even as it begs the question.76
Clear thinking about causality is necessary not only for answering important questions but for asking the right ones in the ªrst place. If by “indictment,” McAllister means “assignment of causal priority,” then I plead guilty
as charged. I lay out a clearly stated causal chain leading back to a rational
public that not only explains the use of a ºawed counterinsurgency strategy,
but also helps address a vexing puzzle that McAllister does not acknowledge, much less solve: Why would a democracy choose to ªght a war it was
unlikely to win at an acceptable cost?77 Despite their disagreements, a larger
truth emerges from McAllister’s article and mine: when a third-party country
ªghts insurgencies, failure is overdetermined. Cost distribution theory suggests a reason why democracies might ªght them all the same.
74. Ibid., pp. 4, 28; and Robert Jervis, “International Politics and Diplomatic History: A Fruitful
Difference,” H-Diplo/ISSF Essays, No. 1 (March 2010), http://www.h-net.org/!diplo/ISSF/
essays/1-Jervis.html.
75. McAllister, “Who Lost Vietnam?” p. 123.
76. Ibid., p. 122.
77. Caverley, “The Myth of Military Myopia,” p. 157.