Cover: Comments on the Wolf-Leites Manuscript

Comments on the Wolf-Leites Manuscript

Rebellion and Authority

Published 1968

by Albert Wohlstetter

In lieu of commenting, as Dan Ellsberg requests, on Ellsberg's comments on the Leites-Wolf manuscript (a possibly infinite regression), I am distributing to his list of persons my own rough comments as of June on the original manuscript.

At Charlie's urging I'm commenting on the manuscript before having read it all through, on the chance that my comments may be more useful now, even though less informed than when I'll be ready for a thorough treatment — some time after your book comes out. Needless to say, I am mostly in agreement with you and think the manuscript will be a very good book. My comments, however, will center on its troubles rather than its felicities.

Apart from incidental comments I will focus on some matters that I gather bother both of you, but especially Nathan: the standpoint of the manuscript with respect to the rebels and with respect to authority. You are right, I think, in feeling uneasy. The manuscript does for a variety of reasons most frequently seem to take the side of the authority. This is something that bothered some of my students at the University of Chicago when they listened to Charlie's account of one of the central chapters; and, in the present climate of partisan extremity, it will identify your book as simply a contribution to the discriminate art of indiscriminate counter-revolution. I think this can be fixed up, and definitely should be, but it may take not only some changes in detail but also some additions and some repair of the central parts of your analytic machinery. Maybe this will emerge from my principal comments.

  1. The sharp distinction made between demand and supply can be useful for analytic purposes but, as you know, is to a considerable extent arbitrary. One always wants to consider in the end choices that are ordered in the light of both costs and benefits. Your analysis explicitly differs from those that concentrate on demand and itself focuses especially on the supply side. It therefore seems to be concerned with means rather than ends (to use everyday language). But especially in connection with rebellions or revolutions this will seem to many of your readers to be precisely the kind of thing that leads governmental bureaucracies like that of the United States astray: an excessive concentration on techniques and an inadequate attempt to relate them to principal short and long-run purposes.
  2. I do not, I would stress, mean that the demand and supply distinction is useless. It can help. However there are ambiguities in your use of it and I think that you need to take greater care on the one hand to relate it to everyday usage and on the other hand to refine and make somewhat more complex the technical analysis. These first remarks and those that follow are directed at the impressions your usage creates in the reader used to everyday language.

  3. The fun you poke at H.A.M. (which I shall comment on at greater length) tends to reinforce the impression that you don't put much importance on what people in a country engaged in civil war actually want. And "importance" is likely to be taken as meaning the same as "value" rather than simply "predictive weight." Of course, in that context you are considering only the conditions for success of an insurgency or counterinsurgency. Not whether either is a good thing. But it reinforces the manipulative impression and it also excludes a number of considerations which many of your readers would regard as paramount. Not only the local population as a whole but the insurgents and the counter-insurgents have a multiplicity of ends and so do the governments outside that support the rebellion or the authority. The appropriateness of the measures taken by R or by A or by their outside supporters can only be measured in terms of the purposes of the various parties. It is plain to a good many Americans today — probably to a good many Vietnamese too — that the end of protecting the Vietnamese population is poorly served by destroying it.
  4. I am not making the banal point here that there is some bystander damage. As Charlie correctly states in answer to my first comments, there always is. More important, an analysis that stresses very heavily making high the price to the population of supporting the rebels and that neglects the preferences of the population for goods and services (or for absolute and relative increases in income) might, in the limit, suggest reducing income per capita so drastically that little would be left over to buy the services of the rebels. (Some other limiting cases are equally dubious.) Defoliation is popularly represented as a move of this kind; and the larger scale disruption of the economy by the process of war may actually be. However the support of people outside (your "X"), to say nothing of the support of the people inside, might be affected by such drastic reductions in population income. Moreover they are bad in themselves and should be considered an item in the calculation. These comments are related to some of the ambiguities in your analysis of demand which I have previously mentioned. And some of which I will comment on in point 3.

    But insurgents have quite analogous problems, problems that anguished radicals in the thirties and forties: the perversion of the revolutionary goals of equality, liberty, etc. by the organizational means used to achieve them. It is particularly necessary now to deal with the relation of the goals to the techniques, or the ends to the means, on each side. All analysis is partial, but I believe you have cut the part too narrowly. The fact (which I have stressed with monotonous frequency in my own writings on systems analysis) that ends can become means to further goals, and that some means are desirable in themselves, doesn't alter the point I am making at all. What I am saying is that the suboptimization you focus on is too low in the hierarchy, given the troubles that both rebels and authorities and their supporters worry about. Needless to say I am not suggesting that the distinction between means and ends is an exciting new discovery. Only that your analysis would be better and less misleading if you more frequently discussed the ends served by insurgency and counter-insurgency.

  5. One of the ambiguities that troubles me affects your use in several major contexts of the notion of "demand." Sometimes you mean by "the demand conditions" to refer to popular preferences for such things as increased dispersion of land ownership, or increased absolute or relative income or the like. So on page III-2 where you refer to demand conditions and say that: " ... dealing with the latter involves the massive problems of modernization in the LDC's. Efforts to overcome these problems are desirable in their own right." H.A.M. advocates (but not only they) are talking about such demands for redressing inequities, etc.
  6. Sometimes, on the other hand, you mean by the "demand conditions" to refer to the demand for R. A very different thing. In a footnote on the same page, for example, you say: "For those interested in conjectural curves, the curve showing 'demand' for R as a function of 'cost' (to P) is likely (perhaps) to be 'kinked'..." And in fact in the text itself immediately following the statement that efforts to overcome these problems "are desirable in their own right" you go on to say: "But the progress that can realistically be aimed for is likely to leave the 'demand' for R fairly strong,..."[1] I am sure you would not in general hold that fulfilling the demand in this second sense, that of conducting a successful rebellion, is necessarily "desirable (in its) own right." You might quite reasonably hold that fulfilling the "demand" in the first sense (that is raising per capita income or the like) is desirable in its own right. Both the authority and the supporters of the authority could conceivably regard an increase in per capita income in the long run as good; and disappointed desires for a successful overthrow of the authority as good too. They might even hold both views and also hold that a short term decrease in the per capita income or some other marked increase in disutility is essential and worth it to defeat the rebellion and to achieve some long-run goals — like increased per capita income. But the double sense in which without distinction you use "demand" here leads to some troublesome confusions. Moreover, my comments may already suggest that this ambiguity conceals the problem of the level of suboptimization by the population, by the rebels, by the authority, and by their supporters. A failure to make the distinction furthers the manipulative impression that I have referred to. I do not think that it is very hard to fix this up. In fact some of my foregoing comments may suggest how it may be fixed simply by making the distinction explicit. And elaborating further some of the complex factors affecting demand and supply for rebellion and for the reductions in inequalities, etc. that are related but not identical with it.

  7. Part of the trouble is that the way economists talk about demand differs from (a) everyday language but also from (b) precise mid-twentieth century mathematical language. It's eighteenth century mathematics. In everyday language when one talks of preferences or choices one doesn't have in mind the static demand function with fixed numbers of consumers, fixed tastes, fixed incomes, and fixed prices for all goods other than the one whose demand is under study. One doesn't mean a static schedule of quantities demanded at various prices. One talks of preferring A to B and, implicit in that preference, are the specific prices of A and B. Supply and demand conditions aren't separated. They are not separated generally when people talk about alternatives in a situation of insurgency. Moreover, so long as one is quite explicit and precise about it, analogous usage in technical studies is quite satisfactory. Systems analyses recommend some systems as preferred over others and express this preference as a result of cost as well as benefit analysis. Moreover one has to be careful in deviating from this popular usage in talking about revolutions. Otherwise, in claiming that what the populace prefers (in the static schedule sense) is much less important than how expensive we can make it for them, one appears to be saying, "Damn what people want." Even when it comes to preferences for rebellion, authorities and their supporters are concerned that the people prefer A to R — all benefits and costs considered.
  8. It is possible to talk more or less precisely of preference both in its everyday form, considering costs and benefits, and in its functional form. On the other hand any one coming to a text in mathematical economics after work in modern metamathematics is likely to be put off somewhat by the looseness of the discussions. Discussions like that of R. G. D. Allen confuse a function variable "f" with "f(x)" which is a variable numerical expression as is the "double of x" which has a constant function and a variable argument. Discussion of parameters, constants, and variable arguments is especially muddled and is one of the things that contribute to the looseness of talk about such functions as demand and supply, each of which is a function of many arguments. Having said this, I feel it seems excessively pedantic, but it is relevant. For in discussing the sorts of demand functions and supply functions you intend in connection with this treatment of rebellion you seem sometimes to fall between two stools — the intuitive everyday stool and the rigorous mathematical one. For rigor you should perhaps explicitly discuss the considerable numbers of arguments one might have to specify for the functions you examine. This would include not only price, individual income, tastes, but also perhaps some of the many factors that are dealt with indiscriminately under the heading of "relative deprivation." And also expected long-term income — absolute, and relative to various terms. There are some clues in the present theory of consumption — Duesenberry, Friedman and others. But at least let me say that I think it would be helpful in clarifying if you discussed some of the uncertainties connected with such arguments [2] to your supply and demand functions. I don't think for example that the income effect is "certain" (II-17), and the preference effect "uncertain." They are both uncertain. The only thing certain about an increase in income is that it increases the possible packages of goods and services purchasable; that is, it is an increase in income. In the short run on a variety of Tocqueville hypotheses it may increase the demands for further changes in income, absolute or relative; and also the demand for the rebels' services. In the long run, as some other Tocqueville hypotheses suggest, it may reduce the demand for revolutionary services, and so on. I have already indicated to you that the Tocqueville hypotheses are more multiple in meaning than you suggest, but I also think it much more uncertain than you suggest when you indicate that the variant espoused by Ed Mitchell has a wide and plain field of application.

    Here again, I do not think it very hard to amend the text. It takes simply a more carefully qualified list of arguments to these functions, confessions of uncertainty, and a recognition that all those other arguments other than price are just like price itself: variables even though they get called "parameters."

  9. On "Hearts and Minds:" I dislike it as much as you and Nathan but have some questions about your treatment. You sometimes suggest that the trouble with H.A.M. is that it considers the preferences of people as decisive or very important. Sometimes you suggest, however, that the trouble is that H.A.M. thinks that majority support is crucial. "As an example of the truism note the following quotation from Hamill: 'The guerrillas ... could not have grown so large (if) most of the peasantry (did not) support them.'" You don't really mean this quote is a truism. You mean that it is false and of course you are correct. Is what's wrong simply a quantitative matter? In the text on II-4 you talk of "... preferential ardor of a large part (for example, 25%, 51%, or 75%) of the population, ..." I think that if the problem of H.A.M. were simply that it overestimated the percent, you should say that. But, in fact, I think the problem is more complicated. I don't have time to go into it, but I do think that the population may in various proportions, some of which are quite relevant (a) be inert, (b) disrupt or otherwise oppose A, (c) favor R, (d) b + c, (e) not simply favor but actively support R or A or both, (f) oppose R, (g) favor A, (h) f + g. I think that some of what you want to say has to do with the small proportion of the population that need and ardently prefer the rebels in order for them to succeed. However, I suspect that some of these other proportions are quite critical and though H.A.M. is wet with sentiment, your own analysis could quite soberly consider these other matters — even under the heading of "demand." In fact, I believe you would do well to represent yourself not as considering supply or emphasizing supply so much as considering a variety of factors on the supply and demand side. I think it is good that you recognize in a footnote on III-18 the interactions between demand and supply. I think you should make more of this.
  10. From here on I am just reproducing what I had written even though the last page or two is garbled.

  11. It is not clear to me how you define "insurgency." You seem not to be dealing with insurgencies in general but to be thinking almost exclusively of Communist rebellions in rural areas of less-developed countries. I am aware that you have several slim references to the MDC, but on III-3 and in many other places you seem to be thinking of the LDC's and also of highly calculating rebels. It would be worth talking today about anarchist types of rebellions and rebellions in the more developed countries.
  12. You have many formulations which suggest that you would like to preserve a balance between the viewpoints of the rebellion and the authority, but you lapse in to the lap of authority. In your chapter VIII you suggest this is "inescapable because a principal concern underlying the book has been improved U.S. policy in insurgent conflicts, regardless of whether it leans in favor of A or R." This seems to me to be wrong on several counts. Among other things it is inconsistent. First of all, it presumes that U.S. policy must favor A. Why? Nothing you have said about insurgency implies that it is bad for the United States or for that matter bad for the local population. Would a successful insurgency against Duvalier necessarily be bad for the United States? Suppose for one thing it was not a Communist insurgency. Or is that excluded by your definition of "insurgency?" Even if it is, while there may be a presumption that a Communist insurgency is bad for the U.S., it conceivably may be less bad than some other alternatives. Or so we thought when we aided the Russians against the Nazis. But in any case, even if insurgencies were always bad for the United States you yourself suggest that some symmetry of treatment is essential; each side needs an understanding of the other, and outsiders need an understanding of both. I believe the passage at the bottom of page 1 and the top of page 2, chapter VIII, is multiply inconsistent. More important, I think that it will not be excessively hard for you to avoid the need for such confused apologies. You can do it with some general statement about your position with respect to the two sides. And with a systematic set of minor revisions of the text of which I'll try to suggest several examples.
  13. I think you should have a general statement which says not that you are "neutral" with respect to insurgencies (as you sometimes protest) but that neither you nor in fact any sensible persons are for insurgency invariably; nor for counterinsurgency invariably; nor invariably neutral. To quote an extract from my own comments on Richard Barnet, a proponent to the New politics: "... whether a revolution is worth supporting or opposing surely should depend on what it is a revolution to as well as what it is a revolution from, and what the human costs are of the revolutionary process, and what other alternatives for change there are ..." Obviously, some rebellions have been the best of a bad lot of alternatives. The Daughters of The American Revolution approve at least one. And no moderately thoughtful revolutionary (not the author of The Permanent Revolution) is in favor of revolution everywhere and forever. Revolutionaries generally want to install a government against which people will not rebel. Except for the anarchists and utopians who hope to eliminate government altogether. But then they think there will be nothing to rebel against.
  14. There is a kind of romantic identification of revolutions as intrinsically good (which you discuss as an American Phenomenon). But this can hardly be a tenable universal position. If you start off therefore by saying that some revolutions might be good and some might be bad and that you are not against all of them or for all of them, that may be useful but not excessively daring. It is useful however since the standard charge made by muddleheads from Fulbright through the New Left is that the U.S. power structure is against revolution any place and will intervene always to defeat them. You are, of course, part of the power structure.

  15. As examples of specific changes to make clear that your book should be of interest to insurgents, counterinsurgents and to their little helpers outside, I have marginalia on:
  16. Prefatory Note p. iii
    p. I–2
    p. I–3
    p. I–5
    p. I–6
    Chapter VIII–1 & 2

    And many other places in the chapters in between that I have looked at. Since Chapter VIII interests you most at the moment, I will illustrate some revisions for it. Chapter VIII, 1 and 2, beginning with the third sentence in the second paragraph, replace the rest of the paragraph with the following:

    In formulating hypotheses and generalizations, we have tried to look at it from the viewpoints of both the Rebellion and the Authority; and from the viewpoints of governments aiding one or the other. Authorities are not invariably worthy of support by those inside or outside the country. Nor are rebels. Nor neutrals. No invariant judgment is possible. If we are successful in the balance we have sought, a hypothetical seminar, consisting of Giap and Magsaysay as representatives of the rebels and the authority, respectively, and T. E. Lawrence and Edward Lansdale as their respective advisers, should find substantial and roughly equal agreement with the propositions advanced in this book.

    I would phrase the statements on page 10 somewhat differently about the effective management of coercion. I would say not that it is unfortunate that coercion is involved but rather that since coercion is involved, it is important for either side that its use by that side be compatible with establishing its legitimacy for A. This means operating under rules of law in order to maintain legitimacy. For R it means using force both to provoke extra legal methods by government and to use force in its own way discriminately so as to suggest that it will displace illegitimate government by a legitimate one.

    This, of course, does not apply to all rebels. For example, it does not apply to anarchist terrorists. And it may not apply to others either. But insofar as a rebel movement intends to replace the existing government with one that will not itself be rebelled against, it is likely to be presenting itself to the population as able to replace the existing disintegrating government with law and order.

    I think you would be wise to talk about the errors of R as well as the errors of A here. The issues have been agitated in radical discussions since the mid-nineteenth century, and Nathan I know is familiar with the literature.

    I have too many additional comments to get them to you this busy week on paper, but perhaps I'd better give you some of the rest orally.


  • [1] See also III–17. "The demand side of the problem relates to what people are willing to pay (or contribute) in order to 'buy' a certain probability or intensity of insurgency."
  • [2] I am or course using "argument" in the mathematical sense.

This report is part of the RAND document series. The RAND Document (D), a product of RAND from 1948 to 1970, was an internal working paper written as a step in a continuing study within RAND, which could be expanded, modified, or withdrawn at any time.

This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.