Afghanistan: Conflict Metrics 2000-2018

By Anthony H. Cordesman

The recent fighting in Afghanistan has shown all too clearly that the Taliban was sincere in announcing in late April that it was rejecting participation in a peace progress and starting a new spring offensive. It has only had limited success so far, but has taken more districts. In spite of a brief Ramadan ceasefire, peace seems no closer now than it has at any point in the past.

One key question is whether the Taliban can make major new gains over the course of 2018. So far, the reporting is mixed. Official reporting by the Government and the ISAF and US command seems overly-optimistic, but Taliban statements also seem overly-ambitious. Equally serious questions also exist over how much of Afghan territory is disputed, controlled by the government, or controlled by the Taliban and other rebels.

Furthermore, there are serious, long-standing serious differences and uncertainties over how to measure the pace of the fighting and each side's level of influence and control. Throughout seventeen years of war, the Afghan government and ISAF have tended to exaggerate their success, "spinning" their news releases to favor their own side. The Taliban has often made its own exaggerated claims, trying to spin the war in the opposite direction.

The Burke Chair at CSIS has again updated a summary historical survey of key combat metrics and official U.S. summary assessments of the war since its beginning to cover the period from January to June 2018. This survey is entitled Afghanistan: Conflict Metrics 2000-2018, and is available on the CSIS web site here.

This survey now traces a long history of official and NGO efforts to portray the course of the war, and provides comparisons of current reporting on government and Taliban success through mid-May 2018– drawing heavily on work by the UN, U.S. government, ISAF, NGOs, and media sources – notably work in the Long War Journal.

Like the original survey, the updated version indicates there is little prospect that a combination of Afghan government, U.S., and allied forces can defeat the Taliban and other insurgent and terrorist forces, or that the Afghan government, U.S., and allied forces will be defeated by them. The conflict has become a war of attrition which can drag on indefinitely, and can only be ended through some form of peace negotiation or the sudden, unexpected collapse of either Afghan government or threat forces – a transition from a war of attrition to a war of exhaustion on one side.

The survey does not provide dramatic new insights into the course of the war: A war of attrition is a war of attrition, but it does warn that the U.S. either failed to properly assess the war from the period after 2003, when the Taliban began to return as a major threat through U.S. plans, to the present. It raises raise serious questions about the combat metrics the United States and its allies have used throughout the war, and the degree that these have been consciously or unconsciously politicized to overstate success or support efforts at withdrawal.

The survey helps illustrate these issues by grouping the data into various time clusters to provide easier comparison. It also provides progressively more competing narratives to help explain what are sometimes major differences in the trends portrayed by given sources. It does not, however, attempt to reconcile the major differences that emerge between sources, or in comparing different types of metrics. In many cases, the source never attempts to defined key terms, indicate the methodology used, or describe the level of uncertainty in the information provided.

In any case, the metrics often do speak for themselves. Anyone familiar with the conflict will be all too well aware of the extent to which the metrics provide in a given period did or did not fully present a valid picture of the war. Anyone who participated in the policies shaping the war over time will be aware of the extent to which official sources chose metrics that exaggerated success, never addressed the deep divisions and lack of actual effective governance on the part of the Afghan government, and emphasize tactical outcomes over insurgent influence.

This update also draws on the work of Bill Roggio and others at the Long War Journal to show just how much estimates of control have changed in the course of 2108. It is clear that major uncertainties exist in current estimates.

There are other several aspects of the survey that the reader should be aware of:

  • Assessments of Taliban vs. government control provided considerable strategic warning from late 2004 onwards that the Taliban were re-emerging as a major threat. The U.S. was slow to react, evidently because the Iraq War had to be given higher priority.
  • The original unclassified assessments by District had six categories, not three. The title changed over time, but reflected a much more realistic picture of the complexity of the situation on the ground: Under government control, largely under government control, contested, largely under threat control, under threat control, and unknown.
  • The graphics, mapping, and reporting on the on the civil side of the fighting, and the effectiveness of the Afghan central government and aid efforts, were largely cancelled after 2011, evidently because the maps and graphics did not reflect the planned level of progress.
  • The estimate of the trends in the number of tactical engagements, and of shifts in them maps of areas under each side's control, sometimes do not seem to track with the UN estimates of the trends in casualties by province and region.
  • The combat metrics highlight the fact that the “surge” in U.S. forces in Afghanistan failed to have a lasting effect and the levels of violence have grown sharply in the process of Transition. A comparison of the previous civil trends, and overall trends in Afghan perceptions, shows the interaction between civil progress and violence, and that the Transition is not succeeding in its current form.
  • Erratic over-classification is a major problem. As SIGAR notes in its April 30, 2018 Report to Congress, there are a number of areas where reporting is not made public where the motive seems to be to spin to war more favorably on downplay serious problems in the Afghan effort.
  • The U.S. data on government or threat control have always been questionable, and sometimes counted Districts as under government control that are actually under the control of various power brokers and warlords, or where the government has only a limited presence in the District capital.
  • The UN data casualty data sometimes seem to reflect an expansion of threat activity that is not reflected in the estimates of control of the disputed districts.
  • Some of the assessments made by governments, the United Nations, media, and think tanks are so different that there is a clear need to improve the official data collection and analysis effort.

The contents are divided into the following major sections:

Title Pages

Watching the Threat Return: Trends and Indicators: 2001-2010 7-23

Failed Surge and Planned Withdrawal: Trends and Indicators: 2011-2014 24-33

Withdrawal Becomes Uncertain and Reality Intervenes: Trends and Indicators: 2015 34-62

Shifting Strategy: Trends and Indicators: 2016 63-91

Conditions Based Strategy: 2017-2018 92-168

Anthony H. Cordesman holds the Arleigh A. Burke chair in strategy at the Center for Strategic and International Studies in Washington, D.C. He has served as a consultant on Afghanistan to the U.S. Department of Defense and the U.S. Department of State.