Even an attempt at a biological or chemical attack would spark panic
Two recent reports about the threat posed by terrorism present contrasting pictures.
The BBC's security correspondent Gordon Correra asks whether terrorism is a real and growing danger, or whether it is in decline?
Three weeks ago, the US released its "Country Reports on Terrorism" for 2007, described as a reference tool for the war on terror.
It highlighted the many plots disrupted over the year around the world, warning that al-Qaeda leaders had "reconstituted", and "continued to plot attacks".
But today, a different view comes from the Human Security Brief for 2007 which comes with the contrasting headline: "Terrorism Fatalities Decline as Muslim Support for al-Qaeda Terror Network Plummets."
"The expert consensus is over-pessimistic," argues the report's author Andrew Mack.
So what is the real picture?
Unfortunately, the answer is that much of it comes down to how you measure and think about the terrorist threat.
Trying to count the number of terrorist incidents and deaths has always been a challenge.
In 2005, the US National Counter Terrorism Center, which is tasked with the job by the US government, changed its methodology, acknowledging that determining whether or not an attack meets any set of criteria is "highly subjective".
Simple numbers of deaths or incidents do not necessarily reflect the reality of a threat and can either be exaggerated or downplayed.
But the report still found a growing number of incidents.
It reported 14,499 terrorist attacks in 2007, half of which led to at least one death (with a total of 22,000 deaths).
Meanwhile, the Human Security Brief argues there has been a 40% decline in fatalities from terrorism.
Are the violent deaths of Iraqi civilians the result of terrorism or civil war?
It does this by challenging the figures used by other counts, particularly when it comes to Iraq, questioning whether violent deaths of civilians in Iraq are really due to terrorism or instead due to a civil war, and if the latter, then why other civil wars - for instance in Sudan or Congo - do not have their fatalities included in counts.
Removing Iraq does makes a significant difference.
The Human Security Brief also argues that if you include civilian casualties, but look at the latter half of 2007, then there is a decline.
The second argument against pessimism from the Human Security Brief is that al-Qaeda is becoming less popular.
There is evidence that in Muslim countries which have been affected by al-Qaeda's terrorism, the organisation has become notably less popular - for instance in Saudi Arabia since 2003.
And in Iraq, Sunni insurgents have turned against al-Qaeda in Iraq.
The brief also points out declining support for attacks on civilians in some Islamic countries and points to polls which show dramatic falls in support for Osama Bin Laden in Pakistan.
This, though, is partly a reaction to the violence that has moved from the tribal areas of Pakistan into the heart of the country.
The reaction may signal declining support for Bin Laden amongst the general population, but it is a product of a much more real terrorist threat.
In a sense this reveals one of the complex problems involved in trying to measure whether terrorism is growing or not.
At exactly the time that terrorism may be a more real threat for people, support for it is likely to decline.
Is that good news or bad news?
Bin Laden's internet messages are getting fewer responses on the web
There is no doubt that in the long run, terrorists need a certain amount of support from populations in order to operate and recruit but in the short run at least, terrorists can continue to do considerable damage whilst being unpopular.
The big question is how "long" is the long term.
But Andrew Mack still argues that all of these trends means that bleak assessments on the terrorist threat and its direction are misplaced.
"When you put all of this together, the impression is of a situation that is better than the expert consensus argues."
There are certainly signs that al-Qaeda's appeal is diminishing to the wider community of Muslims.
Bin Laden's audio and video messages create relatively little chatter on the web forums in the way they used to.
In the long run this trend is bound to reduce the terrorist threat but it may take some time, and the concern is that there are already large numbers of individuals who have already been radicalised, and are still determined to commit acts of violence.
A single major attack like 9/11 could transform the statistics.
And even an attempt to use chemical or biological weapons that do not kill many could create waves of fear, far out of proportion to the statistical toll.
But how likely is that? No-one really knows.
That makes trying to either measure or predict where terrorism is going far from easy.