Home Page
bullet Program for Culture & Conflict Studies Home
bullet About Us
bullet Provincial Overview
bullet Journal
bullet Research
bullet CCS People
bullet Contact Us



Culture& Conflict Studies Banner
Home >>  Culture & Conflict Studies  >>  Welcome
THE CULTURE AND CONFLICT REVIEW Click for an RSS Feed for the Latest Articles

Negotiating with the Taliban: The Timing and Consequences of Settlements in Foreign Power COIN Wars
Andrew J. Enterline and Joseph Magagnoli, 2/1/2011


In early 2009, newly-elected American President Barack Obama undertook a review of the American-led allied war strategy in Afghanistan, Operation Enduring Freedom (OEF). President Obama’s review fulfilled a campaign promise, one grounded in the belief that the Bush Administration failed to implement an effective counterinsurgency (COIN) strategy following a decisive military victory over the Taliban regime in early 2002. This failure to implement an effective COIN strategy resulted in a striking reversal of fortune for OEF by fall 2009, one characterized by a revitalized Taliban insurgency and continued weakness in the Afghan central government (Rohde & Sanger 2007). In fall 2009, President Obama formally endorsed a strategy revision proposed by American General Stanley McChrystal, one grounded in winning the “hearts and minds” of the Afghan population by providing security, economic development, and democracy, a strategy in contrast to its predecessor that was primarily one of searching for and destroying the enemy via the application of military force (McChrystal 2009). Most notably, the cornerstone of this strategy included an infusion, or surge, of 30,000 additional American combat personnel and a 2011 target for commencing the withdrawal American forces (Obama 2009).

Although the option of negotiating with the insurgents emerged in 2008 as the insurgency flourished (UPI 2008), and the feasibility of negotiating with, or “peeling off,” moderate Taliban was considered in early 2009 by President Obama’s policy team (Cooper & Stolberg 2009), the dominant public theme during summer and fall 2009 was one of achieving military victory. Indeed, General McChrystal’s public presentation of his new strategy was anchored to defeating the insurgency, rather than negotiating with the insurgents. Furthermore, following President Obama’s speech outlining the results of his policy review of the war in Afghanistan, American Defense Secretary Robert Gates stated that the United States was in the war “to win.” However, during late fall 2009, references to a revised OEF aimed at securing a negotiated settlement began to emerge in the public media. By January 2010, references to a negotiated settlement suggested that a settlement was the goal of OEF, and that multiple channels of negotiation were being pursued by multiple parties (Press 2010, Rubin 2010, Ward 2009, Gall 2010, Reuters 2010, Wafa 2010).

The emergence of a negotiated settlement as the goal of the American-led allied military mission in Afghanistan raises several questions:

  • How likely is a negotiated settlement with Taliban insurgents?
  • How long will it take to conclude negotiations with the Taliban?
  • What is the likely long-term byproduct of negotiating with the Taliban? How close will the post-settlement “facts on the ground” be to American goals in Afghanistan?
  • How will the recent strategy change in OEF influence negotiations and the resulting short- and long-term consequences?

We investigate these questions by exploring patterns of negotiations between foreign powers and insurgents in COIN wars during the twentieth century. To do so, we elaborate a sample originally reported in Enterline & Magagnoli (2010) to include timing of negotiations between foreign powers, the manner by which foreign powers exit a COIN war, as well as the long-term implications of a negotiated settlement for the issues central to a foreign power’s involvement in a COIN war. Our analysis merely serves as a probe of the aforementioned policy questions, such that we are merely querying the historical record to gain an understanding of how counterinsurgent armies fared in negotiations with insurgents. This probe provides a foundation from which to develop a theory of COIN negotiations that we intend to pursue subsequently.

The remainder of our report is organized as follows. The next section sets out the research design, including our coding of COIN negotiations. The third section executes some descriptive analysis of the data sample on foreign power COIN negotiations during the twentieth century. The fourth section identifies conditions that are central to COIN negotiations historically. We close the report by considering the broader implications of our study for contemporary policy in Afghanistan.

Research Design

As noted, our inquiry into negotiated settlements draws on our previous research (Enterline & Magagnoli 2010), and we refer the reader to this study for an explanation of our approach to studying COIN warfare. Briefly, we examine cases in which foreign power states defend central authority in colonial territories and sovereign states against insurgencies during the twentieth century. We built a sample of 66 cases of foreign power COIN, coding COIN strategy, changes in this strategy, the timing of said changes, and whether the foreign power prevailed militarily over insurgents. For the present inquiry, we code this sample (updated to include an additional case of the Angolan War of Liberation) of 67 twentieth-century foreign power COINs for the type, timing, and long-term impact of foreign power negotiations with insurgents. Some form of negotiations occurs in 52 percent (35/67) of the cases. We detail our codings, which are reported in Table 1, in the remainder of this section.

Negotiation Type

We identify whether foreign powers and/or insurgents sought, or engaged in, negotiations. We identify five manifestations of negotiation in foreign power COIN wars:

  1. Attempt to negotiate by the foreign power. A foreign power fighting a COIN war seeks to engage the insurgents in negotiations to conclude the conflict, but is rebuffed. For example, in the Mozambican Civil War post-coup Portugal sought to negotiate with the FRELIMO insurgents in 1974;
  2. Attempt to negotiate by insurgents. Insurgents fighting a foreign power COIN seeks to engage the insurgents in negotiations to conclude the conflict, but is rebuffed. For example, in the Greek Civil War the Communist insurgents sought to negotiate with the American-supported Greek central government in late 1947 and early 1948;
  3. Collaboration between a foreign power and some element of the insurgency. For example, during the Yugoslav People’s Liberation War the Germany military forces collaborated with the Yugoslav Royalist forces against the Communist Partisan insurgents;
  4. Collaboration between a foreign power and some element of the insurgency, followed by negotiations. A foreign power collaborates with some element of the insurgency and then engages in negotiations with the insurgents. For example, during the Moro Rebellion in the Philippines the United States collaborated with some elements of the Moros while negotiating with remaining combatants; and
  5. Negotiations between the foreign power and insurgents. For example, during the Vietnam (Second Indochina) War, the United States engaged in negotiations with the North Vietnamese and Vietcong insurgents, negotiations that ultimately resulted in the Paris Peace Accords (1973).

The frequency breakdown of the five categories of negotiations between foreign powers and insurgents is reported in Table 2. This tabulation suggests at least two interesting properties regarding negotiations between foreign powers and insurgents during COIN wars. First, collaboration between foreign powers and insurgents is relatively rare, comprising a little over 11 percent (4/35) of the sample. Second, attempts to negotiate are rare, representing about 17 percent (6/35) of the sample, and in only one case (the Greek Civil War) did insurgents seek out negotiations that were refused. Negotiation attempts likely reflect miscalculations by, in most cases, foreign powers regarding the willingness of insurgents to negotiate rather than continue fighting.

Foreign Power Exit

Foreign powers may exit a COIN war in four ways:

  1. Military. The foreign power or insurgents prevail militarily. For example, in the Second Boer War the British defeated the Boer insurgents militarily;
  2. Negotiated Settlement. The foreign power and insurgent forces negotiate the settlement of a COIN war and the departure of the foreign power. For example, following the Cyprus Revolt, the British participated in a negotiated settlement in 1960;
  3. Negotiated Withdrawal. The foreign power and insurgent forces negotiate the withdrawal of the foreign power, but without a broader agreement with respect to broader issues (e.g., the form of the post-COIN government in the host state.) For example, the Egyptians negotiated their withdrawal from the North Yemen Civil War in 1970; and
  4. Withdrawal. The foreign power withdrawals from a COIN war. For example, during the Aden Emergency the British withdrew their forces in 1976 sans neither military victory nor negotiated settlement/withdrawal.

Two forms of foreign power exit from a COIN war—military and withdrawal—reflect instances in which negotiations failed to generate a negotiated outcome for either the foreign power or the insurgents. The two forms of exit that involve negotiations between the foreign power and the insurgents are differentiated, however imprecisely, based on whether the substance of negotiations includes the post-COIN political environment in the host state, rather than strictly about arranging the departure of the foreign power. In our analysis, we treat the two negotiated categories—negotiated settlement and negotiated withdrawal—as one category in which a foreign power exits via negotiation.

The frequency tabulation of this categorization for the 35 cases in which negotiations occurred, were attempted, or collaboration took place is reported in Table 3. A survey of this tabulation suggests that a negotiated exit took place in 60 percent (21/35) of the cases, and as such suggests that foreign powers that engage in negotiations are likely to exit COINs via a negotiated arrangement with insurgents. At the same time, of the 35 cases in which negotiations occurred, were attempted, or collaboration took place, 31 percent (11/35) resulted in a military outcome in which either the foreign power or the insurgents prevailed. 

Long-term Outcome

A central issue with respect to negotiations in Afghanistan concerns the long-term byproduct of such a settlement. Clearly, divining a foreign power’s preferred policy outcome, as well as the long-term outcome of a settlement, are exercises fraught with validity issues. But, the long-term outcome of the war in Afghanistan is central to the acceptability of negotiating with the Taliban—the 64-thousand-dollar question, as it were—and therefore while we acknowledge the difficulty of coding whether the post-COIN outcome is closer to a foreign power or insurgent’s most preferred outcome, we proceed to do so in the spirit of an exploratory exercise.

It is perhaps useful to demonstrate our coding rule for long-term outcome with the contemporary case, one which we assume that the United States and its allies, as well as the insurgents, prefer a long-term outcome that departs significantly from a military defeat. Thus, if American-led coalition negotiates with the Taliban for a power-sharing agreement between the Karzai-led constituents, but the Taliban reneges on this agreement following a departure of the allied COIN force, this turn of events would most likely be judged a policy failure from the perspective of the American-led coalition, but a victory from the perspective of the insurgents. A similar logic is plausible for Taliban engagement in negotiations.

We identify five types of long-term outcomes flowing from COIN wars in which negotiations between foreign powers and insurgents were attempted, collaboration occurred, or negotiations occurred:

  1. Insurgent Defeat. Insurgent forces are defeated or are denied their most preferred policy outcome. For example, during the Cacos War in Haiti, American force negotiated a cease-fire agreement in 1915 that they thought would set the stage for a demilitarization of the Cacos and an end to the insurgency. But, the Cacos resumed fighting and were ultimately defeated by the American military forces;
  2. Insurgent Victory. Insurgent forces achieve most of the policy goals underlying their willingness to engage in an insurgency. For example, in the Angolan War of Liberation, the FRELIMO insurgents prevailed militarily over Portuguese forces. Furthermore, by our coding, insurgents can be victorious, even if they do not prevail militarily. For example, it is generally agreed that during the Algerian War of Independence the French COIN prevailed over the insurgency militarily, but at a political cost that required significant concessions by the French at the negotiating table, the end product of which approximated Algerian insurgent demands that motivated the insurgency in the first place;
  3. Peace Settlement. A foreign power and insurgents agree to a political settlement that does not decisively favor of either combatant. For example, the Troubles in Northern Ireland were concluded in 1998 with the Good Friday Agreement. Of course, any analysis of long-term outcomes with regard to peace settlements is simply provisional, as settlements can fail many years after their agreement. For example, consider that the Paris Peace Accords signed in 1973 to conclude American involvement in the Vietnam war might very well have been coded as a political settlement in 1974, but the eventual collapse of the South Vietnamese government in 1975 following a final North Vietnamese offensive generated an outcome that we argue is distant from the most preferred policy outcome for the United States and qualifies as an insurgent victory; and
  4. Ambiguous. Three cases in our sample—Somalia, the First Chechen War, and Cyprus—reflected long-term outcomes that were ambiguous to the degree such that we felt confident coding them as insurgent victory, insurgent defeat, or peace settlement.

The frequency distribution for our coding of long-term outcome is reported in Table 4, and provides the basis for at least two observations. First, durable peace settlements are rare in foreign power COINs, accounting only about 14 percent (5/35) of the sample. Second, insurgent victory is the modal outcome in the sample, thereby underscoring the fact that historically the odds of a long-term outcome that is closer to the foreign power’s preferred policy position is difficult to achieve.


Our analysis of negotiation focuses on several dimensions of foreign power COIN wars. Specifically, we examine the impact of negotiations on the duration of foreign power COIN wars, the impact of negotiations on the short- and long-term success of foreign powers, the timing of negotiations and long-term outcomes, as well as the relationship between strategy changes, negotiations, and COIN outcomes. We examine each dimensions in turn in the remainder of this section.

COIN Duration

We examine the relationship between negotiations and the the duration of foreign power COIN wars (N=67) during the twentieth century. We find a strong, positive relationship between the two phenomena, such that the decision by a foreign power engage in negotiations (or, in rare cases, be entreated to negotiate) increases the duration to the foreign power’s exit from a COIN war. We illustrate the substantive relationship between negotiation duration in Figure 1. The histogram indicates that negotiations increase the length of foreign power COIN wars by approximately three years above the mean of 6.40 years, with the predicted duration at approximately a decade.[1] It is important to note the relationship between negotiation and the duration of foreign power COIN is unlikely to reflect a causal relationship such that negotiations cause duration entirely. Rather, it is more likely the case that the inability of a foreign power to decisively defeat an insurgency during what is likely to be a lengthy COIN war causes foreign powers to resort to negotiations in order to salvage an outcome that is closest to its ideal point and furthest from military defeat by the insurgents.

Short-term Outcome

We examine the relationship between negotiation and COIN outcomes that we formulated in Enterline & Magagnoli (2010), one that identifies whether a foreign power defeated an insurgency militarily for the full sample of foreign power COIN wars (N=67). In Table 5 we report the frequency tabulation for negotiation and insurgent defeat. Clearly, a strong, negative relationship obtains between the two variables, with negotiations corresponding to a reduced frequency of insurgent defeat. Again, this relationship is unsurprising and likely does not reflect a causal relationship between the two variables, such that negotiations cause a reduction in the likelihood of insurgent defeat. Rather, it is more likely the case that foreign powers that are unable to prevail militarily over insurgent forces resort to negotiations which are rarely designed to achieve military victory.[2] Figure 2 illustrates the substantive reduction in the probability of defeating insurgents when negotiations occur, a change from 81 percent to 37 percent.[3]

Long-term Outcome

As with contemporary Afghanistan, COIN wars in which a foreign power is unable to defeat military its insurgent counterpart demands an answer the following question: Is a negotiated settlement likely to generate a long-term outcome that is close to a foreign power’s preferred policy outcome? As noted, coding the long-term outcome of COIN wars relative to foreign policy preferences on a sample spanning the twentieth century is a difficult, validity-challenged exercise. Yet, it is the paramount policy question for the American-led coalition in Afghanistan, as it pursued General McChyrstal’s (and later General Petraeus’) strategy with the aim of securing a negotiated end to the conflict, but securing a conclusion to the insurgency that is not de facto aligned with Taliban preferences. Table 6 indicates that a foreign power will succeed in achieving long-term success when negotiations occur during a COIN approximately 54 percent of the time, or slightly better than chance.

Duration to Negotiations and Outcome

The occurrence of COIN negotiations is an indication of the balance of power, as well as the prospects for victory, between the foreign power and the insurgents. Seeking negotiations is an indication that one or both combatants are unable to achieve military victory. However, for a number of reasons it is most likely the case that foreign powers are loathe to engage in negotiations with insurgents, particularly early in a COIN war. As such, foreign powers delay negotiating with insurgents under the premise that the insurgency can be broken by continued fighting, or that a strategy change will lead to victory.

In Figure 1 we compute the change in probability of a foreign power experiencing long-term defeat given the timing of negotiations with insurgents at various intervals.[4] Clearly, the predicted probabilities reported in Figure 1 suggest that the longer a foreign power waits to negotiate with insurgents, the lower the probability that the foreign power will prevail in terms of the long-term, post-COIN outcome. Indeed, the probability of a foreign power experiencing long-term defeat is 46 percent if negotiations occur in year five of the COIN war, yet the probability of defeat jumps to 67 percent if a foreign power waits until year 15.

The relationship between the timing of negotiations and long-term outcome reflects what we refer to as a foreign power’s negotiation paradox. The negotiation paradox in COIN wars is a phenomenon in which foreign powers delay negotiating in the hope that military victory can be achieved. Yet, by delaying negotiations for victory, or even doing so to increase leverage in negotiations, the probability that the foreign power will achieve long-term success shrinks. This dynamic likely reflects a vicious cycle: The longer a foreign power waits to negotiate in lieu of military victory, the less likely that negotiations will yield long-term outcomes that are close to the foreign power’s most preferred outcome.

Strategy Change

The recent transition to a population–centric COIN policy, implemented after General McChrystal’s appointment as Commander of NATO forces in Afghanistan, adds an additional layer of complexity to the negotiations process. As such we address two issues pertinent to Afghanistan. First, we examine the extent to which a strategy change affects the ability of a foreign power to gain a negotiated settlement such that the foreign power can withdraw. Further, we identify ideal points in the conflict in which a settlement and subsequent foreign power withdrawal is most likely. Last, we examine, the probability that a foreign power will gain a successful long-term settlement for the foreign power after a strategy change.

Negotiated Withdrawal

Figure 4 reports a plot of the probability of negotiations that enable the foreign power to withdraw after a strategy change.[5] The plot reflects the curvilinear pattern in which the probability of a negotiated withdrawal is very low but increases overtime until the probability declines again. However, the overall probability of a negotiated withdrawal after a strategy change is very low. The reason for the low probability of a negotiated withdrawal is that foreign powers shift strategy in order to decisively defeat the insurgency, rather than pursue a negotiated outcome.

However, if the United States is seeking a negotiated end to the current COIN war in Afghanistan, the question becomes at what point after a strategy change will negotiations result in a settlement? Figure 5 displays the probability of achieving a negotiated withdrawal for different durations between a strategy change and the first year of negotiations.[6] The probabilities displayed in Figure 5 suggest that there is a low probability of achieving a negotiated withdrawal relatively soon after a strategy change. The curvilinear relationship suggests that the probability of achieving a negotiated withdrawal is maximized (a 70 percent chance) when negotiations take place approximately 3.5 to 4 years after the foreign power changed strategies. Thus, while negotiated agreements are possible, the chances of gaining an agreement very soon after a strategy change are probabilistically slim.

Long Term Outcome

While negotiated agreements are possible after a strategy change, the real question concerns the effectiveness of these agreements in the long term? Figure 6 displays the probability of long term success following a strategy change.[7] The graph displays several notable points. First, immediately following a strategy change, the probability of gaining an outcome favorable to the foreign power in the long term is approximately 50 percent. This is not surprising because in many cases when a foreign power must change its strategy during the conflict, the course of the conflict is most likely not favoring the foreign power.

Second, as a COIN war continues the probability of a foreign power gaining a favorable long-term settlement actually drops precipitously for the first four years after a strategy change. At approximately 4.5 to 5 years following a strategy change, the probability of achieving a favorable long term settlement begins to increase. The probability of a successful long-term outcome is maximized at approximately nine years after a strategy change, with a 70 percent chance of successful long term settlement. This trend makes sense because strategy changes, such as the switch to a population based COIN in contemporary Afghanistan, take time and commitment to implement. A strategy shift can reflect a long term commitment towards the foreign power’s policy goals, and in turn a willingness to continue fighting. Yet, foreign powers that engage in strategy switches followed by rapid negotiations to end a conflict shortly most likely signal a lack of commitment, thereby increasing insurgent recalcitrance. As a result, a sharp decline in the probability of successful long term outcome by a foreign power is observed shortly after a strategy change.

Discussion and Implications for Afghanistan

Much like their insurgent counterparts, foreign powers fight COIN wars to win. During the twentieth century foreign powers prevailed 58% (39/67) of the time, but this rate of success shrinks to 46 percent (17/37) in the post-1945 period. Failing to defeat insurgents militarily means that foreign powers seek to negotiate with insurgents in order to secure policy goals that are preferred to outright military defeat. Thus, few foreign powers will pursue a negotiated settlement when outright military victory is attainable; rather, foreign powers resort to negotiations when military victory is unattainable.

The resort to negotiations presents several issues for combatants in general, and foreign powers in particular. First and foremost, by engaging in negotiations a foreign power acknowledges, by deed if not word, that it is unable to defeat the insurgents militarily. Second, the decision by foreign powers to resort to negotiations often reflects a rising intolerance of a foreign power’s domestic audience to bear the additional costs associated with continued prosecution of the war. Last, COIN negotiations present the problem of absent guarantor, such that any settlement between a foreign power and insurgents is difficult to enforce following the foreign power’s departure.

These issues reinforce the reluctance of foreign powers to engage in negotiations with insurgents, a reluctance that we call the negotiator’s paradox. The negotiator’s paradox reflects the reluctance of foreign powers to negotiate early in a COIN war when insurgent willingness and capacity to fight and defeat the foreign power are unknown to the foreign power and vice versa. The primary behavior highlighted in negotiator’s paradox—foreign powers postponing negotiations with insurgents—is exacerbated by the fact that the support of domestic audiences in foreign powers for continued prosecution of COIN wars is a negative function of time, and is likely to approach its nadir at the very point when a foreign power resorts to negotiation.

Indeed, the erosion of public support spurs, in part, the willingness of a foreign power to pursue negotiations. This erosion of public support often manifests itself in demands by domestic audiences for extrication of the foreign power from the COIN war. This extrication imperative further erodes a foreign power’s leverage over a negotiated settlement and, worse, increases insurgent confidence that capitulation by the foreign power is possible, and therefore the potential for insurgents to achieve an outcome that is very close that which would be achieved by military victory over the foreign power. Thus, once the value of extrication exceeds a foreign power’s willingness to continue fighting, the leverage that a foreign power can exert on a negotiated settlement is significantly reduced.

As noted, an additional issue for foreign powers, one that further retards a foreign power’s willingness to negotiate with insurgents, is the absence of a capable and/or willing third party guarantor of any settlement that emerges between the foreign power and the insurgents. At the same time, the absence of a capable third party makes a negotiated settlement attractive for insurgents, because in the post-settlement period insurgents can push the “reality on the ground” toward an outcome that is closer, or even achieves, the insurgents most preferred outcome of the war and do so without sanction by a third party.

The absence of a capable and willing guarantor is a function of several forces. First, the pool of capable guarantors in world politics is small. Second, the inability of the negotiating foreign power to defeat the insurgents reveals to would-be guarantors how difficult and costly it would be to enforce a guarantee. Last, in the context of world politics, foreign power COIN wars are central to strategic relationships between powerful regional and global actors. In short, one state’s failure in a COIN war is another state’s advantage. Thus, the pool of capable states that are sympathetic to securing a foreign power’s negotiated settlement is further constrained.

When combined, the extrication imperative and the absence of a guarantor generate additional complexities that make it difficult for the foreign power to achieve post-settlement success. For example, once it is known that a foreign power wishes to exit a COIN war short of achieving military victory, parties to the conflict, ranging from the central government that the foreign power seeks to defend, to insurgents, to the benefactors of the insurgents, may defy coordination necessary for a negotiated settlement, as the interests of these parties are realigned given the anticipated departure of the foreign power.

For example, the central government that the foreign power is fighting to protect may reject a negotiated settlement between the foreign power and insurgents because this unguaranteed outcome will leave the central government exposed to insurgent demands, the most extreme of which is political or physical annihilation. As a result, the extrication imperative may provide an incentive for the foreign power-backed central government to seek new benefactors, scuttle foreign power negotiations, or conclude a separate settlement with the insurgents. Thus, the very point in COIN wars at which foreign powers are inclined to negotiate presents very difficult conditions for achieving an acceptable settlement are not conducive to such an outcome.

Thus, foreign powers confront the vexing problem of extracting themselves from a costly conflict in which military victory is unattainable, while attaining a negotiated settlement that approximates their goals for engaging in a COIN war in the first place. The problem is that these two goals are contradictory, because the insurgents are aware of the costs the foreign power is absorbing and can deduce that the foreign power wants to extract itself from the conflict. Furthermore, continued intervention is very costly to the foreign power, which must project and supply its forces and often considerable distances, but relatively less costly for the insurgents, which are locally based.[8] This disparity in the costs of fighting manifests itself in a greater bargaining leverage for the insurgents at the negotiating table.

Thus, in order for the foreign power to achieve an agreement that meets its own criteria it finds itself compelled to increase its own bargaining leverage before negotiating with insurgents. A foreign power achieves this increase in leverage on the battlefield. In Afghanistan, this effort to increase foreign power leverage is manifest in the ongoing operation in Marjah in which American/NATO personnel are attempting to wrest southern Afghanistan from Taliban control. However, efforts by foreign powers to use short-term gains on the battlefield as foundations for favorable long-term settlements are undermined by the absence of a capable third party guarantor. Indeed, when insurgent concessions flow from short-term increases in leverage achieved by a foreign power, insurgents have more to gain later by reneging on this agreement after the foreign power departs.

By compelling deeper compromises by the insurgents through short-term escalation, foreign powers are building in incentives for long run failure into the agreement because the negotiation is not consistent with insurgents view of the conflict. Rather, short-term increases in pressure by foreign powers can create equilibria that artificially increase a foreign power’s leverage during negotiations with insurgents. This dynamic is evident in Figure 6. If foreign powers seek to settle the conflict quickly after a strategy change, the probability of long term success drops to approximately 10 percent. The trend reverses the longer the duration after a strategy change. To avoid failure in the long term the foreign power must continue to act as a security guarantor in the long term. The probability of successful long-term settlement increases to approximately 70 percent after nine years following a strategy change.

The implications of this analysis are worrisome for the Allied COIN war in contemporary Afghanistan. Without a significant and credible commitment of time and resources by Allied forces, a successful long-term policy in Afghanistan is unlikely. Domestic and economic pressure demands that the Allied force depart rapidly. Yet, the Allied force prefers to maximize its negotiating leverage in any settlement to which it is a party, so a strategy change and subsequent military offensives are underway to increase leverage.

We argue that this newly acquired leverage is derived from an unsustainable equilibrium, one that is subject to reversal by Taliban insurgents that are convinced that Allied forces are too irresolute to maintain said equilibrium over the medium- or long-term. Even if the Taliban is incorrect and the Allies are resolute in their commitment to maintaining this equilibrium, or if the equilibrium achieved by the Allies via military gains is indeed very costly to the Taliban, the Taliban may negotiate with the knowledge that the presence of a capable guarantor following Allied departure is unlikely. This dynamic provides an incentive for the Taliban to seize power from a Karzai-led central government that lacks indigenous support and capacity after the Allied force draws down below levels adequate for settlement enforcement.

We argue that the evolution of the American-led OEF in Afghanistan reflects the hallmarks of the negotiator’s paradox in foreign power COIN wars. Foreign powers seek to win militarily, and negotiate only when winning is impossible, which may be due to military, political, and economic constraints that emerge individually or in combination. This dynamic means that foreign powers pursue negotiations later rather than sooner in COIN wars, and in doing so negotiate at the point when its leverage and credibility in negotiations are most subject to erosion, thereby undermining long-term success.

About the Authors

Andrew J. Enterline is an Associate Professor of Political Science, specializing in international relations, at the University of North Texas. His research interests include studies of democratization and interstate conflict, imposed political systems, and third party involvement in civil wars.

Joseph Magagnoli is a Doctoral Student in international relations in the Department of Political Science at the University of North Texas. His research interests include counterinsurgency, military strategy, and quantitative methodology. The authors are grateful to the Department of Political Science and the College of Arts and Sciences at the University of North Texas for general research support. The contents herein are the authors’ and do not represent the views of the University of North Texas, its Board of Regents, or the State of Texas.


Associated Press (2010). “Aide: Karzai ‘Very Angry’ at Taliban Boss’ Arrest.” New York Times, March 15, 2009.

Associated Press (2009). “Gates to Afghans: US ‘in This Thing to Win’.” New York Times, December 8, 2009.

Cooper, Helene & Sheryl Gay Stolberg (2009). “Obama Ponders Outreach to Elements of Taliban.” New York Times, March 8, 2009

Drew, Christopher (2009). “High Costs Weigh on Troop Debate for Afghan War.” New York Times, November 14, 2009.

Enterline, Andrew J. and Joseph Magagnoli (2010). “Reversal of Fortune? Strategy Change & Counterinsurgency Success by Foreign Powers in the Twentieth Century.” University of North Texas, Department of Political Science, Typescript. http://www.psci.unt.edu/enterline/em-freshlook-v58-identified.pdf.

Gall, Carlotta (2009). “As U.S. Weighs Taliban Negotiations, Afghans Are Already Talking.” New York Times, March 11, 2009.

McChrystal, Stanley (2009). “Commander’s Initial Assessment.” NATO International Security Assistance Force, Afghanistan. Accessed August 30, 2009. Redacted (Unclassified) Version.

Obama, Barack (2009). “Obama’s Address on the War in Afghanistan.” New York Times, December 2, 2009.

Reuters (2010). “U.N. Set For More Discreet Talks With Taliban.” New York Times, March 15, 2010.

Rohde, David and David E. Sanger (2007). “How a Good War in Afghanistan Went Bad.” New York Times, August 12, 2007.

Rubin, Alissa J. (2010). “Arrests of Top Taliban Figures Ended Talks, Ex-Envoy Says.” New York Times, March 19, 2010.

UPI (2008). “Momentum builds for Taliban negotiations.” UPI.com, October 13, 2008.

Wafa, Abdul Waheed (2010). “Departing U.N. Envoy Urges Political Solution in Afghanistan.” New York Times, March 4, 2010.

Ward, Jon (2009). “Biden: Taliban negotiations likely.” Washington Times, March 11, 2009.1.


1. Figure 1 is based on a bivariate OLS regression. To control for outlying observations (primarily very long COIN wars) in our sample, we re-estimated the model with a natural log of foreign power duration. In turn, we employed the predlog routine to generate a smear value that enables a substantive interpretation of the regression coefficient of the logged duration. The smear value of duration is 2.4 years, which is very close to the unlogged coefficient of approximately 3 years and overall substantive story suggested by the analysis.

2. That said, in some cases—e.g., East Timor—it can be argued, at least retrospectively, that combatants negotiate for time in order to resume fighting in an effort to secure a military victory.

3. Our logit estimation is of form π=α+β1X1, generating the following estimates that we employ to compute predicted probabilities (standard errors in parentheses): ^π=1.47(.456)−1.99(.576)X1.

4. Our logit estimation is of form π=α+β1X1, generating the following estimates that we employ to compute predicted probabilities (standard errors in parentheses): ^π=−1.465(.812)+.808(.447)X1.

5. Probabilities are based on restricted cubic spline regression estimates, which are non–parametric estimates based on a smoothed binary response variable. The number knots used as anchors for the spline are chosen by the number that minimizes the Akaiki Information Criterion (AIC). 3 knots were used for Figure 4 and the probabilities are robust to different knot selections.

6. Probabilities are estimates from a restricted spline regression. The AIC is minimized using 3 knots, and the probabilities do not change substantively with different numbers of knots. Negative values on the x-axis reflect negotiations that occurred before a strategy change, while a zero value indicates that a strategy change that occurred in the same year as negotiations, and positive values indicating the number of years following a strategy change in which negotiations took place.

7. Probabilities are estimates from restricted cubic spline regression with 4 knots. Four knots were chosen because they minimize the AIC. However, the probabilities change substantively when using 3 knots. The biggest difference between the two models is that when using 3 knots the rebound in probability of long term success is tempered, displaying only an increase from a low of a 28% chance of success at 7 years to a mere 33% chance at the 17 year mark.

8. In November 2009, the United States estimated that it cost “about $1 million per soldier a year” to fight in in Afghanistan (Drew 2009).  



Material contained herein is made available for the purpose of peer review and discussion and does not necessarily reflect the views of the Department of the Navy or the Department of Defense.
The appearance of external hyperlinks does not constitute endorsement by the United States Department of Defense, the United States Department of the Navy and the Naval Postgraduate School of the linked web sites, or the information, products or services contained therein.