Wednesday, April 22, 2009
Why Don't All Terrorists Claim Credit? We Have An Answer.
Earlier I blogged some pretty speculative ideas about why some terrorists do not claim credit for their attacks. I now learn that Bruce Hoffman also thinks this is an interesting question and that he beat me to the punch by a significant margin, publishing an article on the topic in Terrorism and Political Violence back in the spring of 1997. Check it out if you're interested in the issue--his ideas are far better than mine.
Does Torture Stop Terror? Nope.
Torture is in the news. Bush-era memos justifying the use of violence against detainees have just been released. The Bush people that wrote or acted on these memos justify such violence in the name of stopping terrorism. The claim is that the violence led the detainees to give up information that allowed the authorities to foil terrorist attacks in the planning stages.
As many have pointed out in recent days, it's difficult to know with any certainty that this is the case. It's possible that the detainees would have given up the intel even if gentler techniques of questioning had been used (as professional interrogators for the FBI and other agencies have long claimed). It's also possible that the authorities would have found about the impending attacks from other sources, or that the attacks would have been called of for any number of reasons. So the debate is stuck in a circle, with supporters of the methods claiming they worked, and opponents saying that they did not.
One partial solution is to look at different types of data. In particular, do countries that practice torture heavily experience less terrorism? The answer is pretty clearly "no." Here's how I arrived at this conclusion:
My colleague Jim Piazza and I have a paper coming out in Comparative Political Studies that analyzes the influence of human rights abuses on terrorism. We find that, for a wide range of data sources, control variables, and statistical specifications that governments that abuse rights actually experience more terrorism.
But what about torture? Torture is only one of the broader range of human rights we looked at in the paper. Is the relationship of torture to terror different? It might be, since far more countries engage in torture than in other forms of violent human rights abuses. To answer this question, I re-analyzed the data from the paper. Details are in the following paragraph; skip down if you just want the punchline.
I re-estimated the three models described in the paper that use the MIPT measure of terrorism as the dependent variable. This counts the number of terrorist attacks in each country from 1998 to 2004 committed by domestic and transnational groups, and also combines these into a measure of all terrorism. I used a negative binomial regression with robust standard errors clustered on countries and the same independent variables as those reported in the paper (political participation, constraints in the executive, regime durability, international war, civil war, and the logs of population and GDP per capita). I replaced the independent variable measuring human rights with two new variables. The first is a measure of torture from the CIRI project. The second is the measure of human rights used in the paper minus torture. This is meant to capture the possibility that torture and other human rights abuses are substitutes for each other; a regime might not torture, say, but could still have a bad record of respecting other rights.
Torture has a negative and statistically significant relationship to terrorism in all three models. In other words, countries that engage in more torture (and thus have a lower score on the torture variable) consistenly experience more, not less, of both domestic and transnational terrorism. This mirrors the more general finding reported in the paper that respect for human rights is associated with less terror as well.
What are the implications for the debate in the US today? The clearest is that torture does not work, at least in reducing terorism. It's another nail in the coffin for those who justify torture as a tool of counterterrorism. It also suggests that we don't need to worry about how revealing the details of the US torture program will provide terrorists with the skills to avoid providing information to interrogators. Intead, it suggests that a suprisingly easy and morally unambiguous counterterrorism strategy is to be nice to people. Being mean (like, say, torturing) seems to annoy some victims, who go on to become or serve as examples to new terrorists.
As many have pointed out in recent days, it's difficult to know with any certainty that this is the case. It's possible that the detainees would have given up the intel even if gentler techniques of questioning had been used (as professional interrogators for the FBI and other agencies have long claimed). It's also possible that the authorities would have found about the impending attacks from other sources, or that the attacks would have been called of for any number of reasons. So the debate is stuck in a circle, with supporters of the methods claiming they worked, and opponents saying that they did not.
One partial solution is to look at different types of data. In particular, do countries that practice torture heavily experience less terrorism? The answer is pretty clearly "no." Here's how I arrived at this conclusion:
My colleague Jim Piazza and I have a paper coming out in Comparative Political Studies that analyzes the influence of human rights abuses on terrorism. We find that, for a wide range of data sources, control variables, and statistical specifications that governments that abuse rights actually experience more terrorism.
But what about torture? Torture is only one of the broader range of human rights we looked at in the paper. Is the relationship of torture to terror different? It might be, since far more countries engage in torture than in other forms of violent human rights abuses. To answer this question, I re-analyzed the data from the paper. Details are in the following paragraph; skip down if you just want the punchline.
I re-estimated the three models described in the paper that use the MIPT measure of terrorism as the dependent variable. This counts the number of terrorist attacks in each country from 1998 to 2004 committed by domestic and transnational groups, and also combines these into a measure of all terrorism. I used a negative binomial regression with robust standard errors clustered on countries and the same independent variables as those reported in the paper (political participation, constraints in the executive, regime durability, international war, civil war, and the logs of population and GDP per capita). I replaced the independent variable measuring human rights with two new variables. The first is a measure of torture from the CIRI project. The second is the measure of human rights used in the paper minus torture. This is meant to capture the possibility that torture and other human rights abuses are substitutes for each other; a regime might not torture, say, but could still have a bad record of respecting other rights.
Torture has a negative and statistically significant relationship to terrorism in all three models. In other words, countries that engage in more torture (and thus have a lower score on the torture variable) consistenly experience more, not less, of both domestic and transnational terrorism. This mirrors the more general finding reported in the paper that respect for human rights is associated with less terror as well.
What are the implications for the debate in the US today? The clearest is that torture does not work, at least in reducing terorism. It's another nail in the coffin for those who justify torture as a tool of counterterrorism. It also suggests that we don't need to worry about how revealing the details of the US torture program will provide terrorists with the skills to avoid providing information to interrogators. Intead, it suggests that a suprisingly easy and morally unambiguous counterterrorism strategy is to be nice to people. Being mean (like, say, torturing) seems to annoy some victims, who go on to become or serve as examples to new terrorists.
Thursday, April 16, 2009
Message to Joseph Nye: I Am Relevant
Joseph Nye published an op-ed in the Washington Post a few days ago arguing that academic political scientists are, well, too academic. We don't reward those among us who engage in policy debates. Policy relevance is at best ignored and at worst punished. This has attracted a lot of attention in the small world of political science blogging--see, for example, Peter Howard, Daniel Drezner, and Henry Farrell.
OK, but what does "policy relevance" mean? Us academics have this habit of insisting on definitions. Nye does not provide one. He implies that an academic is relevant when he or she does work that can contribute to contemporary policy debates. But where's the line between policy relevance and irrelevance?
I cannot find it. Consider Theda Skocpol's States and Social Revolutions. This book--enormously influential in political science if only because every graduate student, even the Americanists, have to read it at some point--analyzes the causes of revolutions in France, Russia, China, and other countries (some of which don't exist, like Prussia, so are not very policy relevant). A nice piece of academic work, perhaps, but it would seem to be totally irrelevant for DC types, right? But aren't these types thinking about the possibility of revolution in, say, Pakistan? And weren't their predecessors criticized for ignoring revolutions in Iran and in the Soviet Union?
So maybe studying revolutions could be a bit "relevant", and not just interesting for ivory tower types. Indeed, the same Theda Skocpol that generated dozens of pages on Prussia applied her ideas to the sources of the Iranian revolution shortly after it occured. Wouldn't this have had the potential to help policymakers understand what what happening in Iran in the 1980s? Oh, and the CIA in the 1990s funded a project--run by academics--to analyze the sources of revolution. I doubt they did this because they just thought revolutions were, um, interesting.
Policy relevance is important. But do we really want the academy stuffed with people who are up to date on the latest developments in country x? No. That's what think tanks and CIAs are for. We (by which I mean I, of course) want an academy with people that are interested in the bigger causes and consequences, and are trained to think about these things in a systematic way. How come? One of the best ways that academics can be policy relevant is to poke holes in dumb ideas. Every academic I know (bar one, and he was Canadian of all things) thought the invasion of Iraq was a dumb idea. They knew this based on their knowledge of the history of preventive war, positive theories about preventive war, and long thought about the normative implications of the use of force to make the world a better place. This is all stuff that gets downplayed in "policy circles," but if it was taken more seriously might at least prevent us from making the worst mistakes over and over again.
OK, but what does "policy relevance" mean? Us academics have this habit of insisting on definitions. Nye does not provide one. He implies that an academic is relevant when he or she does work that can contribute to contemporary policy debates. But where's the line between policy relevance and irrelevance?
I cannot find it. Consider Theda Skocpol's States and Social Revolutions. This book--enormously influential in political science if only because every graduate student, even the Americanists, have to read it at some point--analyzes the causes of revolutions in France, Russia, China, and other countries (some of which don't exist, like Prussia, so are not very policy relevant). A nice piece of academic work, perhaps, but it would seem to be totally irrelevant for DC types, right? But aren't these types thinking about the possibility of revolution in, say, Pakistan? And weren't their predecessors criticized for ignoring revolutions in Iran and in the Soviet Union?
So maybe studying revolutions could be a bit "relevant", and not just interesting for ivory tower types. Indeed, the same Theda Skocpol that generated dozens of pages on Prussia applied her ideas to the sources of the Iranian revolution shortly after it occured. Wouldn't this have had the potential to help policymakers understand what what happening in Iran in the 1980s? Oh, and the CIA in the 1990s funded a project--run by academics--to analyze the sources of revolution. I doubt they did this because they just thought revolutions were, um, interesting.
Policy relevance is important. But do we really want the academy stuffed with people who are up to date on the latest developments in country x? No. That's what think tanks and CIAs are for. We (by which I mean I, of course) want an academy with people that are interested in the bigger causes and consequences, and are trained to think about these things in a systematic way. How come? One of the best ways that academics can be policy relevant is to poke holes in dumb ideas. Every academic I know (bar one, and he was Canadian of all things) thought the invasion of Iraq was a dumb idea. They knew this based on their knowledge of the history of preventive war, positive theories about preventive war, and long thought about the normative implications of the use of force to make the world a better place. This is all stuff that gets downplayed in "policy circles," but if it was taken more seriously might at least prevent us from making the worst mistakes over and over again.
Wednesday, April 15, 2009
Intelligence Sharing: Comparing the US and the EU
The Markle Foundation has issued another in a series of reports on how the United States can improve information sharing among its intelligence agencies and senior policymakers.
Improved intelligence sharing was, of course, a major objective of the intelligence reforms instituted since 9/11. The most prominent of these reforms was the creation of the Office of Director of National Intelligence. Many argued that the creation of the ODNI was a shell game, since the Director had little authority to change agencies' budgets, reassign personnel, and so on.
Implicit in this criticism is the assumption that governments operate as effective hierarchies, with leaders (ODNI) ordering subordinates (CIA, NSA, etc.) around. Of course this is sometimes the case. But it's not a realistic model for the intelligence community. Many intelligence agencies have corporate cultures that resist outside interference, and others are parts of cabinet agencies that themselves do not answer to the ODNI (State, Energy, etc.), and most of the rest are part of the Department of Defense and thus can draw on its bureaucratic and budgetary influence to ignore a meddling ODNI.
The Markle report is interesting becuase it recognizes this reality. Instead of arguing that the ODNI needs much more authority to carry out its' goals, the report suggests more modest changes that would encourage each intelligence agency to see sharing as in its interest. These include technological changes that would make sharing easier, the expansion of social networks that facilitate collaboration by lower-level intelligence people across agency lines, keeping intelligence sharing decentralized so no agency feels like the others can take advantage of it, and encourage common training regimes across agencies.
This agenda is certainly less ambitious than, say, pretending to again reorganizing the intelligence community from top to bottom. A more hierarchical system might work in some contexts (i.e. the military seems to do pretty well with it outside of intelligence), but our intelligence community is too decentralized and has interests and cultures that are too divergent to make it cost-effective. The Markle report recognizes that such reorganizations often do not address the root causes of failure, and impose costs in terms of time, money, and attention.
Indeed I argue in a paper I'm presenting next week at the meeting of the European Union Studies Association that similiar changes could improve intelligence sharing in the European Union. Sure, the US is a single country and the EU is not. So the barriers to a hierarchical intelligence sharing system in the European Union are likely higher than they are in the US. Nonetheless, the dynamics are remarkably similiar--the national intelligence services in the EU, like agencies in the US intelligence community, do not fully trust each other to treat intelligence securely, worry that they would send more intelligence than they would recieve, and fear a central authority that does not understand their corporate culture and technical needs.
Improved intelligence sharing was, of course, a major objective of the intelligence reforms instituted since 9/11. The most prominent of these reforms was the creation of the Office of Director of National Intelligence. Many argued that the creation of the ODNI was a shell game, since the Director had little authority to change agencies' budgets, reassign personnel, and so on.
Implicit in this criticism is the assumption that governments operate as effective hierarchies, with leaders (ODNI) ordering subordinates (CIA, NSA, etc.) around. Of course this is sometimes the case. But it's not a realistic model for the intelligence community. Many intelligence agencies have corporate cultures that resist outside interference, and others are parts of cabinet agencies that themselves do not answer to the ODNI (State, Energy, etc.), and most of the rest are part of the Department of Defense and thus can draw on its bureaucratic and budgetary influence to ignore a meddling ODNI.
The Markle report is interesting becuase it recognizes this reality. Instead of arguing that the ODNI needs much more authority to carry out its' goals, the report suggests more modest changes that would encourage each intelligence agency to see sharing as in its interest. These include technological changes that would make sharing easier, the expansion of social networks that facilitate collaboration by lower-level intelligence people across agency lines, keeping intelligence sharing decentralized so no agency feels like the others can take advantage of it, and encourage common training regimes across agencies.
This agenda is certainly less ambitious than, say, pretending to again reorganizing the intelligence community from top to bottom. A more hierarchical system might work in some contexts (i.e. the military seems to do pretty well with it outside of intelligence), but our intelligence community is too decentralized and has interests and cultures that are too divergent to make it cost-effective. The Markle report recognizes that such reorganizations often do not address the root causes of failure, and impose costs in terms of time, money, and attention.
Indeed I argue in a paper I'm presenting next week at the meeting of the European Union Studies Association that similiar changes could improve intelligence sharing in the European Union. Sure, the US is a single country and the EU is not. So the barriers to a hierarchical intelligence sharing system in the European Union are likely higher than they are in the US. Nonetheless, the dynamics are remarkably similiar--the national intelligence services in the EU, like agencies in the US intelligence community, do not fully trust each other to treat intelligence securely, worry that they would send more intelligence than they would recieve, and fear a central authority that does not understand their corporate culture and technical needs.
Tuesday, April 7, 2009
AfPak Stategy: A contradiction in terms?
This is troubling. It's a summary of a memo written to troops in Afghanistan by General David D. McKiernan, commander of the U.S. and NATO forces in the country.
Here are the punchlines:
No one expects that there will ever be many more than 100,000 western troops in Afghanistan. We know from earlier conflicts, including those analyzed in James Dobbins' book on the topic, that this is far too few to undertake an effective counterinsurgency campaign along the lines that McKiernan suggests. And it's not even clear that a military-led counterinsurgency campaign with sufficient resources can achieve all the objectives laid out here, including fighting the enemy, promoting development, and also developing a functioning local government. After all, in most cases where such state building has been successful, the builder did not also have to fight a major insurgency (think of west Germany or Japan). And in many cases without an insurgency the results have not been that great (think of Kosovo).
Here are the punchlines:
- The focus should be on "governance, development and security concurrently" because "success in Afghanistan will not come from the sole pursuit of a security line of operations by military forces."
- "Do not clear an area unless GIRoA [Government of the Islamic Republic of Afghanistan] and the ANSF [Afghan National Security Forces] are able to hold it."
- "facilitate solutions . . . that reinforce the rule of law and GIRoA's legitimacy," taking care "not to strengthen local powerbrokers working outside governance structures."
No one expects that there will ever be many more than 100,000 western troops in Afghanistan. We know from earlier conflicts, including those analyzed in James Dobbins' book on the topic, that this is far too few to undertake an effective counterinsurgency campaign along the lines that McKiernan suggests. And it's not even clear that a military-led counterinsurgency campaign with sufficient resources can achieve all the objectives laid out here, including fighting the enemy, promoting development, and also developing a functioning local government. After all, in most cases where such state building has been successful, the builder did not also have to fight a major insurgency (think of west Germany or Japan). And in many cases without an insurgency the results have not been that great (think of Kosovo).
Monday, April 6, 2009
Why Don't All Terrorists Brag?
Brian Jenkins semi-famously wrote decades ago that "terrorists want a lot of people watching, not a lot of people dead." The second part of this has been questioned by work on the "new" terrorists, which holds that terrorist do want to wreak mass destruction.
But what about the first part? Convention wisdom holds that terrorism is a political tactic, not a military one. Terrorists are too weak to inflict much direct harm on their opponents. Instead they use violence to communicate through their message to supporters, opponents, and fence-sitters.
If the paramount motive is publicity, why is it so difficult to identify the attackers in many cases? Why, for example, Al Qaeda often not claim responsibility for attacks it carries out? Why is it that so many attacks summarized in databases such as the Global Terrorist Database, Iterate, and WITS cannot be attributed to particular groups?
This puzzle raises important questions about how we think about terrorism. If terrorists are really not that into communication, what does that say about their motives and strategies? What else do they want?
Let me suggest some answers. Some are pretty straightforward, such as measurement error. But others raise interesting possibilities for re-thinking the strategic calculus that underlie terrorist attacks. In particular, they might lead us to think more seriously about the varied audiences to which terrorists appeal, and how they use different techniques to reach each audience:
1. Some terrorists are dumb, and don't understand the importance of communication. So they don't call the TV stations after blowing something up. This is certainly possible, but probably won't take our understandingo of the logic of credit claiming very far since, as Stephen Krasner pointed out some time ago, "stupidity is not a very interesting analytical category".
2. Some are just into it for the violence, and political demands are just a justification for this.
3. My colleague Jim Piazza points out that many unattributed attacks cause few or no causalities. Perhaps terrorists see these as failed attacks, and don't want to claim credit for them.
4. Piazza also notes that many are in countries that experience a large number of attacks, so it might be attributable to measurement error brought on by the chaos of frequent terrorism. For example, a bank robbery may be counted as a criminal act in a country with little terrorism, but as a terrorist attack in a country where real terrorists regularly rob banks.
5. My other colleague John Szmer is an American politics scholar, and suggest that some groups may have as their true audience not the general public for specific supporters or potential supporters. They might be able to communicate their responsiblity for an attack to such supporters directly, rather than through the mass media.
6. Different people have different definitions of terrorism. This includes the terrorists themselves. Say terrorists attack a military outpost. The victims might see this as "terrorism". But the perpretrators might view it as legitimate military action, for which they do not want to brag to the newspapers.
7. It might depend on the political goals or ideology of the group. Religious terrorist groups, for example, face the problem that most major religions frown on killing innocent civilians. So a religiously-motivated terrorist group could engage in attacks, but not claim all of them so as to avoid alienating some co-religionists.
Evaluating many of these logics (and others you might think of) raises that always tricky chicken and egg problem. It's difficult to explain why terrorist group x did not claim credit for its attack, since we don't know that it was terrorist group x that committed the attack, since terrorist group x did not claim credit for the attack.
It might be possible, though, to develop other observable implications for each rationale for not claiming an attack. To take a simple example of this, point 4 above suggest that countries with lots of political violence have more unattributed attacks. Point 7 suggests, perhaps, that countries characterized by much religious terrorism would have more unattributed attacks. Point 5 might imply that the larger social network of a terrorist group influences its credit-claiming strategy, so countries with terrorist groups that are able to communicate with audiences in ways other than the media might have more attacks.
But what about the first part? Convention wisdom holds that terrorism is a political tactic, not a military one. Terrorists are too weak to inflict much direct harm on their opponents. Instead they use violence to communicate through their message to supporters, opponents, and fence-sitters.
If the paramount motive is publicity, why is it so difficult to identify the attackers in many cases? Why, for example, Al Qaeda often not claim responsibility for attacks it carries out? Why is it that so many attacks summarized in databases such as the Global Terrorist Database, Iterate, and WITS cannot be attributed to particular groups?
This puzzle raises important questions about how we think about terrorism. If terrorists are really not that into communication, what does that say about their motives and strategies? What else do they want?
Let me suggest some answers. Some are pretty straightforward, such as measurement error. But others raise interesting possibilities for re-thinking the strategic calculus that underlie terrorist attacks. In particular, they might lead us to think more seriously about the varied audiences to which terrorists appeal, and how they use different techniques to reach each audience:
1. Some terrorists are dumb, and don't understand the importance of communication. So they don't call the TV stations after blowing something up. This is certainly possible, but probably won't take our understandingo of the logic of credit claiming very far since, as Stephen Krasner pointed out some time ago, "stupidity is not a very interesting analytical category".
2. Some are just into it for the violence, and political demands are just a justification for this.
3. My colleague Jim Piazza points out that many unattributed attacks cause few or no causalities. Perhaps terrorists see these as failed attacks, and don't want to claim credit for them.
4. Piazza also notes that many are in countries that experience a large number of attacks, so it might be attributable to measurement error brought on by the chaos of frequent terrorism. For example, a bank robbery may be counted as a criminal act in a country with little terrorism, but as a terrorist attack in a country where real terrorists regularly rob banks.
5. My other colleague John Szmer is an American politics scholar, and suggest that some groups may have as their true audience not the general public for specific supporters or potential supporters. They might be able to communicate their responsiblity for an attack to such supporters directly, rather than through the mass media.
6. Different people have different definitions of terrorism. This includes the terrorists themselves. Say terrorists attack a military outpost. The victims might see this as "terrorism". But the perpretrators might view it as legitimate military action, for which they do not want to brag to the newspapers.
7. It might depend on the political goals or ideology of the group. Religious terrorist groups, for example, face the problem that most major religions frown on killing innocent civilians. So a religiously-motivated terrorist group could engage in attacks, but not claim all of them so as to avoid alienating some co-religionists.
Evaluating many of these logics (and others you might think of) raises that always tricky chicken and egg problem. It's difficult to explain why terrorist group x did not claim credit for its attack, since we don't know that it was terrorist group x that committed the attack, since terrorist group x did not claim credit for the attack.
It might be possible, though, to develop other observable implications for each rationale for not claiming an attack. To take a simple example of this, point 4 above suggest that countries with lots of political violence have more unattributed attacks. Point 7 suggests, perhaps, that countries characterized by much religious terrorism would have more unattributed attacks. Point 5 might imply that the larger social network of a terrorist group influences its credit-claiming strategy, so countries with terrorist groups that are able to communicate with audiences in ways other than the media might have more attacks.
Wednesday, April 1, 2009
Books End
The University of Michigan Press plans to release most new books in electronic format. This makes a lot of sense. Kindle, netbooks, and Iphones give us more and better ways to read digital text. And I cannot remember the last time I read a journal article in a physical journal--instead it's much more convenient to download the article from JSTOR or some other database. And the Press saves the costs of paper, ink, etc. so might be able to produce more or cheaper books (or "books" if you thing this is a bad idea).
If I had a manuscript in preparation, though, I'm not sure I would be very excited about this. In many fields, publishing a book is essentially a requirement for tenure and promotion. But academics are very conservative about things like this, and I suspect many would devalue a book/"book" that is not published in the traditional format. Michigan will review these books the same way it always has, so actual quality should not decline, but percieved quality might.
Eventually books (and journals) printed in the traditional manner will go the way of the 8 track, and in the long run presses and authors can benefit from this. But there is a real coordination problem involved in getting from here to there. Many presses and authors are reluctant to be the first to move in this direction because they want to maintain percieved quality. Once some of the more prestigious presses begin to do so, though, the others may follow quite quickly. If Cambridge and especially Princeton (at least in political science) moved in this direction, they could mitigate some of the concerns about the quality of electronic books, since they have a very strong reputation for producing must-read scholarly work. Their shift to electronic publishing could quickly wipe out the quality issue, and lead to a cascade of other presses following their lead.
If I had a manuscript in preparation, though, I'm not sure I would be very excited about this. In many fields, publishing a book is essentially a requirement for tenure and promotion. But academics are very conservative about things like this, and I suspect many would devalue a book/"book" that is not published in the traditional format. Michigan will review these books the same way it always has, so actual quality should not decline, but percieved quality might.
Eventually books (and journals) printed in the traditional manner will go the way of the 8 track, and in the long run presses and authors can benefit from this. But there is a real coordination problem involved in getting from here to there. Many presses and authors are reluctant to be the first to move in this direction because they want to maintain percieved quality. Once some of the more prestigious presses begin to do so, though, the others may follow quite quickly. If Cambridge and especially Princeton (at least in political science) moved in this direction, they could mitigate some of the concerns about the quality of electronic books, since they have a very strong reputation for producing must-read scholarly work. Their shift to electronic publishing could quickly wipe out the quality issue, and lead to a cascade of other presses following their lead.
Subscribe to:
Posts (Atom)