In August 2006 I received a Guestbook question about how to manage terrorism. In my answer, I discussed the complex relationship between managing terrorism itself and managing people’s fear of terrorism. I quoted extensively from a recent exchange of emails with terrorism policy expert Frank Harvey, focusing on Harvey’s hypothesis that successful counter-terrorism programs actually increase the public’s threat perception. When I sent Harvey a draft of my Guestbook answer, he responded, and I included parts of his response in the revised answer.
A more complete version of my email exchange with Harvey is below.
From Frank Harvey to Peter M. Sandman (June 7, 2006)
You might be interested in a research report I just completed for the Canadian Defence and Foreign Affairs Institute….
I’d love to get your thoughts on the main argument, which I think is directly relevant to the events surrounding the counter-terrorist “success” in Toronto over the past week.
The Homeland Security Dilemma will be the focus of my research when I take up the 2007 J. William Fulbright Research Chair in Canadian Studies at SUNY (Plattsburgh) in January.
From Peter M. Sandman to Frank Harvey (June 11, 2006)
I enjoyed your report on The Homeland Security Dilemma. Thanks for sending it to me.
I have no quarrels with your use of my work, or of the psychometric approach generally.
I do have some reservations about your overall thesis. I agree with your basic point – that successful security leads the public to expect that security will be successful. Evidence of success at any goal leads to expectations of continued success. Parents of kids who get good grades expect good grades. Residents of neighborhoods with a low crime rate expect a low crime rate. The record frames expectations for the future. So I don’t doubt that people expect a higher level of success in deterring terrorism than they expected in early 2002. Whether they expect it because of how much money has been spent or because of the absence of successful attacks inside North America is hard to assess. My guess would be that the record of success is the main driver, though surely the expenditures would be the main evidence of grievance if/when the effort fails.
- Are people really more fearful of terrorist attack? People claim to be worried, but does it affect behavior? Do we spend less time in crowded public places, for example? I’d say we have achieved the New Normal, more fearful than in August 2001, no doubt, but significantly less fearful than in October 2001.
- Are people really more likely to overreact to a successful attack? Are you confident that a second 9/11 would generate a more powerful emotional reaction than the first? Or even that it would lead to more scapegoating than the first? My intuition is that it would not – that people would be more saddened than frightened, and would say they’d known all along our luck couldn’t hold. What about a tinier success on American soil – an attack that killed two or three people? You predict a huge overreaction. I’m not so sure. Even if there were a huge overreaction the first time, what about the third time?
- How might people’s level of fearfulness and expectations of success be affected by a government communication effort that stressed, simultaneously, the good success record that has been achieved and the impossibility of sustaining it? My sense is that the U.S. government has actually stressed neither as much as I would have recommended. Claims of past success are, perhaps, constrained by security concerns themselves, which lead to suppression of the details that would give such claims more drama and more credibility. And claims of future failure are constrained by a fear that the public will reward the politician who promises the impossible more than the politician who manages expectations realistically. So we end up with a government that seems to be making an implied promise of future perfection, but keeps it deniable by not making it explicitly and keeps it unbelievable by not documenting the sources of past success. This seems pretty unwise to me. I would be saying, very explicitly: We’ve done a lot better than we ever expected. We expect to continue to do well … but not THAT well. Like a team with a perfect season so far, we are bound to lose a game eventually, though we surely expect to win the title. (This is the wrong metaphor to use, but I imagine the point is clear.)
- Can we distinguish impacts on expectations from impacts on emotions? Safer roads lead people to expect fewer traffic accidents – but do they really lead people to fear traffic accidents more? A better comparison: Reduced crime leads people to expect fewer crimes, but does it lead them to fear crime more? Somewhere in the middle between expectations and emotions are post-failure recriminations – which are, as you note, a big piece of what motivates politicians. Do we punish a police commissioner more if the crime rate goes way down and then part-way back up than if it stays steady or declines slowly? Does it help if the police commissioner says last year was terrific and we probably won’t do that well this year, but we’re pretty confident we won’t go back to the bad old days either?
- Your collection of the survey data is impressive. But I’m not sure confidence that another attack can be prevented is the right measure; improved acceptance of reality would mean reduced confidence (but without recriminations). Approval of the effort to prevent an attack is a better measure – but “approval of the War on Terror” is too contaminated with attitudes toward the Iraq invasion. Are there any questions that ask more narrowly about approval of the effort to prevent an attack on U.S. (or Canadian) soil? Better yet, I’d love to track a question like “How many Americans/Canadians do you expect will be killed by foreign terrorists inside the country in the next X years?” I’ll bet the number has gone down since the months after 9/11, though it’s doubtless still higher than before 9/11.
- It is presumably good news that people expect fewer domestic deaths from foreign terrorism than they expected right after 9/11. The downside of the good news is that higher expectations lead to more criticism of the occasional failure, which leads to more security expenditures in pursuit of perfect success. That’s the fundamental truth in your article. Lower expectations, though, would lead to depression (emotional and perhaps economic), and also to increased demand for improvements, and thus for increased security expenditures. If higher expectations and lower expectations both lead in the same direction, the curve has to reverse direction in the middle somewhere! Figuring out the shape of this curve sounds like a worthwhile task to pursue. So does figuring out how communication efforts can shift the curve.
- After five years of perfect success, the first domestic attack that isn’t thwarted will no doubt have a big effect – even if government communicators do more to predict that this will inevitably happen (but not often). But what about the second, or the third? Are the inflated expectation problems provoked by unsustainable success self-correcting when the inevitable failures occur? Or do you predict that a successful attack would/will lead to more demands for perfection, more expenditures aimed at achieving perfection, and more implied promises to achieve it? Once again, a curve that goes down at both ends has to have a top. What frequency of attack success yields the most realistic public?
- How much of your thesis is true in a country that has had considerable but not perfect success thwarting terrorists. What are the data, for example, from Israel? Or from Northern Ireland during the relevant decades?
Well, enough of that.…
From Frank Harvey to Peter M. Sandman (June 14, 2006)
Thanks so much for taking the time to provide such excellent assessment of the CDFAI paper and thesis. You may recall we had a similar exchange following the publication of another one of my papers in the International Journal (2003) – “Addicted to Security: Globalized Terrorism and the Inevitability of American Unilateralism.” …
I will no doubt send along a more detailed response to your observations at some point in the future, as I work on completing my next book on the Homeland Security Dilemma, but my immediate reaction is similar to the one I provided in our last exchange. We simply have a different opinion about the stage the U.S. is currently in. I agree with you that the public, at some point in the future, will grow accustomed to the threat and risks and achieve a “new normal,” as you put it. But we are nowhere near that point yet. Using Israel as the benchmark for expectations and predictions, we are essentially at 1948.
If the last five years are any indication of a typical U.S. reaction to terrorism, and if this reaction applies in a relatively benign security environment (with an almost perfect homeland security record), imagine what will happen after the next attack. As I point out in the paper, the problem is not the “failure of imagination” – we’ve pretty well nailed that one. The problem now is the “imagination of failure.”
Consider the following. As a result of the Toronto area arrests of 17 terrorist suspects last week, 70% of Canadians now believe there will be other terrorist attacks in Canada in the near future. In other words, what may turn out to be one of the most successful counter-terrorist operations in history has “increased” threat perceptions in Canada to U.S. levels. Politicians will be very reluctant to question those perceptions, and that’s the dilemma….
From Peter M. Sandman to Frank Harvey (June 14, 2006)
Isn’t it rational to deduce from the arrest of 17 terrorist suspects in Toronto that terrorists are targeting Canada (or if they were gathering to attack the U.S., that they might target Canada now)?
If I had previously thought a successful attack on Canada was unlikely because Canada had such excellent counterterrorism programs, then the recent news should have reduced my probability-of-successful-attack estimate. But if I had previously thought a successful attack was unlikely because terrorists have bigger fish to fry, then the news should have increased my probability-of-successful-attack estimate.
If you parse the probability of successful attack into the expected frequency of efforts to attack and the expected frequency of success in those efforts (that is, expected failure frequency of Canadian counterterrorism), then the recent arrests are bad news vis-à-vis the first factor and good news vis-à-vis the second.
For me the goal is a public that expects periodic terrorist efforts, a good but not perfect counterterrorism program, and thus occasional successful attacks – a public that supports the counterterrorism effort; that weighs the pros and cons of more elaborate counterterrorism efforts (financial cost, civil liberties cost, likely payoff in improved prevention); that ends up supporting some but not all “improvements”; and that accepts an occasional successful attack as preferable to a bankrupt police state (but also accepts some financial and civil liberties cost as preferable to an open invitation to attack). The policy question is how to achieve that sort of public.
As you say, we’re both looking for ways to move toward the New Normal – but we have different hypotheses about where we are on the path.
From Frank Harvey to Peter M. Sandman (August 13, 2006)
PS: I have periodically exchanged emails on this topic with Canadian security expert Frank Harvey. Harvey argues that people, at least in the developed countries of the west, are becoming “addicted to security.” He describes what he calls “the homeland security dilemma” – that successful security efforts lead people to feel more vulnerable and demand still more security.
FH: Not exactly – the homeland security dilemma (HSD) is a much broader thesis that encompasses predictions about public and government behavior/perceptions. Simply put, the greater the financial costs, public sacrifice and political capital invested in security, the higher the public’s expectations and corresponding standards for measuring performance, the more significant the public’s sense of in-security after each failure, and, paradoxically, the higher the pressure on governments and citizens to sacrifice even more to achieve perfect security. The central argument can be summed up by the following counterintuitive thesis: The more security you have, the more security you will need. This is not because enhancing security makes terrorism more likely (although the incentive for terrorists to attack may increase as extremists feel duty-bound to demonstrate their ongoing relevance), and it is not because successful security measures lead people to feel more vulnerable. It is because huge investments in security (and other financial, political and civil liberty sacrifices) raise public expectations and demands, so that even small failures and perhaps even a few major successes (like those in Toronto and London) increase threat perceptions and amplify public outrage. Moreover, the public’s perceptions (whatever they may be) are not the only issue here – political assumptions about public perceptions determine public policy. If leaders expect a public backlash after small failures, then they will do as much as they can to reach perfection – even if that benchmark is impossible and irrational. Newly implemented security rules regarding prohibiting liquids on flights are a case in point.
PS: In June 2006, Harvey wrote to me: Consider the following – as a result of the Toronto area arrests of 17 terrorist suspects last week, 70% of Canadians now believe there will be other terrorist attacks in Canada in the near future. In other words, what may turn out to be one of the most successful counter-terrorist operations in history has “increased” threat perceptions in Canada to US levels. Politicians will be very reluctant to question those perceptions, and that’s the dilemma.
FH: This is accurate, Peter.
PS: I’m not convinced that Harvey is right. I think it is very difficult for people to sustain their fear of terrorism, or indeed of any risk. A successful terrorist attack like 9/11 generates an enormous amount of fear for a while, but soon enough – too soon, many would say – most of us return to the “new normal,” more fearful of terrorism than before 9/11, but less fearful by far than right after 9/11.
FH: Yes, but, as I have noted in other email, we disagree about where we are with respect to the “new normal,” and the ripple effects of a superpower’s response to real (or imagined) attacks will continue to be system-transforming. Moreover, the public may be less fearful than they were immediately after 9/11, but they will be much angrier today than they were on 9/11 if a repeat of 9/11 occurs today. That is the reality political officials face, and the paradoxical consequence of spending billions to prevent another 9/11. I also believe that the public’s fear of terrorism will be significantly higher today than it was on 9/11 if a repeat of 9/11 occurs.
PS: Is the U.S. and European public excessively worried about terrorism today, as Harvey seems to believe? Or is it insufficiently worried? (Or, conceivably, does the public have it about right?)
FH: One way to answer the question is to compare the level of fear (typically derived from perceptions of the risks) with the actual probability of being a target/casualty. 30–50% of Americans (and now Canadians) are worried about being a target, and a much higher percentage (50–70%) believes another attack will take place, despite the very low probability of these events actually occurring, and despite the perfect homeland security record to date.
PS: I do agree that successful security efforts have a paradoxical impact. They are experienced in part as near-misses, and thus as evidence of vulnerability. Media coverage of the recent arrests in the U.K., for example, routinely described them as bad news; reporters saw the story less as civilization’s victory over terrorism than as proof that the terrorists are out there. And Harvey is also right that successful security efforts make people inclined to expect that future security efforts will be successful.
FH: The other problem with “successes” like Toronto and London is that they confirm the existence of domestic threats, homegrown hatreds, and a clear intention to kill. It will always be easier to re-define any success as a “failure,” and that is the challenge political officials face when trying to establish more balance in public perceptions. The other problem is that meaningful successes are really hard to prove (more recent evidence about the London plot is that the attack was not imminent). When you add the very rational “political” interests of opposition groups in Canada and the U.S. to exploit government incompetence (or to question “successes”) the problem gets even more complex. You may be right, Peter, about what the government should be doing to manage public risk perceptions, but there are simply too many constraints and pressures that I believe will prevent them from following those recommendations. The problem is that this option is an equally rational public policy path to take, which explains why officials face a dilemma – a choice between rational options that produce negative, costly and unintended consequences.
PS: This may or may not lead people to demand more security. But it almost certainly leads people to see any future terrorist success as a security failure deserving of blame. It’s easy to imagine the investigations, recriminations, and scapegoating that would follow a second 9/11. And it’s easy to understand why government officials are tempted to respond to this thought experiment with additional layers of security. As Harvey has put it, “the problem is not the ‘failure of imagination’ – we’ve pretty well nailed that one. The problem now is the ‘imagination of failure.’”
PS: The risk communication lesson here, I think, is the urgent need to communicate that terrorism risk management isn’t dichotomous – that while a perfect record is certainly our goal, it isn’t our expectation and shouldn’t be the public’s expectation. Here’s how I put it in an email to Harvey: How might people’s level of fearfulness and expectations of success be affected by a government communication effort that stressed, simultaneously, the good success record that has been achieved and the impossibility of sustaining it?
FH: But consider the public backlash if officials followed your recommendations, spent more time highlighting (perhaps even exaggerated) successes, and then suffered another small failure. The public would have reason to suspect complacency if not arrogance and the backlash would arguably be much worse. This is certainly how the issue would be exploited by opposition groups, and why so little time is spent on successes.
PS: My sense is that the U.S. government has actually stressed neither as much as I would have recommended. Claims of past success are, perhaps, constrained by security concerns themselves, which lead to suppression of the details that would give such claims more drama and more credibility.
FH: Right, but that is a difficult problem to resolve.
PS: And claims of future failure are constrained by a fear that the public will reward the politician who promises the impossible more than the politician who manages expectations realistically.
PS: So we end up with a government that seems to making an implied promise of future perfection, but keeps it deniable by not making it explicitly and keeps it unbelievable by not documenting the sources of past success. This seems pretty unwise to me. I would be saying, very explicitly: We’ve done a lot better than we ever expected. We expect to continue to do well … but not THAT well.
FH: It may be unwise, but it’s inevitable – and that’s the dilemma.
Copyright © 2006 by Frank Harvey and Peter M. Sandman