The Argument from Crisis and Pessimism Bias
Many people have argued that the public seems to have an overly negative view of society’s development. For instance, this survey shows that the British public think that the crime rate has gone up, even though it has gone down. Similarly, Hans Rosling points out that the public has an overly negative view of developing world progress.
If we have such a pessimism bias, what might explain it? One standard explanation is that good news isn’t news—only bad news is. A murder or a famine is news; their absense isn’t. Hence people listening to the news gets a skewed picture of the world.
No doubt there is something to that. In this post I want, however, to point to another mechanism that gives rise to a pessimism bias, namely the compound effect of many uses of what I call the Argument from Crisis. (Please notify me if you’ve seen this idea somewhere else.)
The Argument from Crisis says that some social problem—say crime, poverty, inequality, etc—has worsened and that we therefore need to do something about it. This way of arguing is effective primarily because we are loss averse—because we think losing is worse than failing to win. By arguing that inequality was not as bad ten years ago and that we have now “lost” some degree of equality, your argument will be rhetorically stronger. The reason is that in that case more equality will eradicate a loss, whereas if inequality hasn’t worsened, more equality will simply be a gain, which we value less. Hence we will be more inclined to act against inequality in the former case.
Even though the distinction between a gain and an eradication of a loss is important from a rhetorical point of view, it does not seem very relevant from a logical point of view. Whatever the level of crime or inequality is, it would seem that the value of reducing it is the same regardless of whether it has gone up or down the past ten years.
Another reason for why the Argument from Crisis is rhetorically effective is of course that we believe that whatever trend there is will continue (rightly or wrongly). Hence if we think that crime or inequality is increasing, we believe that it will continue do so unless we do something about it.
Both of these factors make the Argument from Crisis rhetorically effective. For this reason, many people argue that social problems which they want to alleviate are getting worse, even though in fact they are not.
I’d say the vast majority of people who use this argument are not conscious of doing it, but rather persuade themselves into believing that the problem they want to alleviate is getting worse. Indeed, I think that the subconscious use of this argument is a major reason why radicals often think the world is on a downward slope. The standard view is of course that they want radical change because they believe that the world has got worse, but I think that to some extent, the causality is reversed: they believe that the world has got worse because they want radical change.
Since the Argument from Crisis is so rhetorically effective, it gets used a lot. The effect of this is to create, among the public at large, a pessimism bias—an impression that the world is getting worse rather than better, in face of evidence to the contrary. This in turn helps various backward-looking political movements. Hence I think that we should do more to combat the Argument from Crisis, even though it can sometimes be a rhetorically effective means to persuade people to take action on important social problems.
There is a rational difference between “crimes (or inequality) has gone up and we need to do something” and “crimes (or inequality) has gone down, but we still need to gain more” : diminishing returns/feasibility issues.
If crimes (or inequality) was indeed lower 10 years ago, it means that reaching a lower point in crime/inequality was possible with a reasonable cost. Of course, it might be impossible now (because of a change in technology, in the environment, in whatever external things we don’t really control), but the fact that it was lower before is valid evidence (even if not proof) that is possible to get it lower with a reasonable effort.
On the other hand, if crime or inequality is at an absolute low, it’s harder to know if it’s possible to get it even lower, or how much effort that would require.
Of course, that doesn’t change the core of your reasoning, but it does make the “Argument from Crisis” more than just a rhetorical device—at least, when it’s used accurately, ie, when the things are actually getting worse.
Good. I actually gave one reason that shows that there is a rational difference between “crimes (or inequality) has gone up and we need to do something” and “crimes (or inequality) has gone down, but we still need to gain more”:
“Another reason for why the Argument from Crisis is rhetorically effective is of course that we believe that whatever trend there is will continue (rightly or wrongly). Hence if we think that crime or inequality is increasing, we believe that it will continue do so unless we do something about it.”
You point to another important such reason, however.
Still my hunch is that loss aversion is a major reason why the Argument from Crisis is so effective.
Measuring crime rates isn’t easy. If citizens don’t trust the police to go after criminals they report less crimes. The article you linked to along with many other source therefore use the murder rate as a proxy because presumably most murders get reported to the police. That’s problematic because better emergency medicine decreased the death rate among people who get shot.
It’s stronger than that. Over time newsrooms in a lot of countries have made the decision to spent more time reporting crimes.
Naomi Klein’s book “The Shock Doctrine” might be relevant.
There’s Rahm Emanuel with “You never let a serious crisis go to waste. And what I mean by that it’s an opportunity to do things you think you could not do before.”
Robert Higgs wrote “Crisis and Leviathan”, chronicling the growth of government power through crises. People who want government to grow have an interest in creating actual or perceived crises.
The media is overwhelmingly populated by the party of ever bigger government. And they coincidentally peddle crisis after crisis.
Took me a while to connect the dots. You’re arguing that the supposed decrease in crime is actually an artifact of decreased reporting to police and increased effectiveness in hospital trauma care? Wouldn’t attempted murder rates go a long way toward clearing that up?
The debt ceiling is used as a crisis to shrink government. Regardless of your goals, if you want political change in our Western democratic culture a crisis is useful.
I think that it likely that over the last decades overall crime rates went down. On the other hand some of the change in crime numbers we see is likely through other effects and whenever you have the debate about the subject you should be aware of those factors.
You should always be a bit skeptic that the numbers that you have actually describe reality fully.
If you count crimes, counting murder rates has the advantage that if there’s a shot dead body, it’s quite obvious that a person was killed. There might be some issue with separating manslaughter from murder, but a dead body nearly always triggers procedures in the official bureaucracy.
The same is not true with attempted murder. In a community where the police isn’t trusted by the population some attempted murders won’t be reported. That’s why arguments such as the one linked in the OP are often made via the murder rate.
We do have another tool for determining the amount of crime: Victimization surveys. But they also aren’t perfect.
They’ve tried again and again and again. Government gotten smaller yet?
The debt ceiling itself is not a tactic of “crisis”. It is an attempt to put limits on government spending. Leaving it to a “crisis” is the successful tactic of those who would and have blown through it time and time again.
I believe the republicans are in full capitulation on this point. Maybe with both chambers of congress they’ll feel that they can control the narrative by putting a bill on the president’s desk for him to sign or not. But I doubt it.
Yeah, but I would think people are pretty motivated to report actual attempts to kill them.
I certainly agree with that. Conscious of abstration, etc.
I hadn’t thought about Klein and Emanuel’s ideas, though they do seem somewhat related. Good!
I think the key difference is that you made the argument here about mechanism that convince individuals while Klein and Emanuel focus on systematic feature of our Western democracy.
I would expect the strategy of talking in terms of crisis to be a lot less effective in China to create political change than it is in the West and thus less used.
If you want to imply that the rates may not fall as clearly as indicated that is probably not going to work. Almost all counts are pointing straight down—except for steps upward when the rules for counting cases are changed. The data for Germany is publicly available (apparently reported more or less since the 1950s) and provides lots of indicators to look at:
German Crime Statistics since 1999 (English)
dito detailed cases since 1987 (German)
You ignored the second sentence.
What’s tracked in your links is cases. That means crimes that come to police attention. Crimes that come to police attention isn’t the same thing as crimes.
If you put more policemen on the streets the amount of crimes goes down but the amount of crimes that come to police attention goes up.
Programs for raising awareness of domestic violence makes victims come forward and report crimes to the police. That raises official crime rates but there no good reason to assume that it raises the actual rate of crimes.
I’m not sure what you are driving at. Reported and actual rates are clearly shown in my linked sourced and overall both go down over a wide range of incidents except for apparent jumps when case counting rules change.
“Actual rates” is not something that we now and can list in a statistic. We can only use different methods to get a proxy of that value.
Yes yes. Any measurement has an error. I shouldn’t have written “actual”. But what are you driving at? Do or don’t you agree?
You have written actual because you think “actual”. It’s not.
If you take your first link and actually read the paper you will find: “The unlawful (criminal) acts dealt with by the police, including attempts subject to punishment, are recorded in the Police Crime Statistics. This also includes the drug offences handled by the customs authorities. ”
Police Crime Statistics is the name of the report. It doesn’t include crimes not dealt with by the police because nobody reported the crime and the police has no other way of knowing about it.
It’s not like we don’t know anything about the dark figures. The first rigorous approach seems to be this one:
Dunkelfeldforschung in Bochum 1986⁄87 by Hans-Dieter Schwind
This is in German, but an executive summary in English can be found on page 290 ff.
We know enough to know that statistics like the one you cited shouldn’t be treated as being the actual values and that was the main point I was making above.
You have a good case that pessimistic rhetoric is effective, but I’d hesitate to conclude that pessimistic bias is primarily caused by exposure to such rhetoric.
It’s an interesting question how to test this. Do people in controlled-media states consistently think things are getting worse? Do hunter-gatherers?
Yeah, I’m not sure about that either. But it strikes me as one possible mechanism among others that could cause pessimism bias.
Another is this: suppose that the world is getting better in some ways, and worse in others. If you are loss-averse, you’re going to weigh the losses more heavily than the gains. This means that you will think that the world is getting worse even though it is not for someone who is not loss-averse (i.e. biased against gains).
OK, but note that the most reliable crime statistics also come from a survey. In fact, the same survey! So rather than saying that this is because people are too negative, you have to tell a more subtle story.
What we really have is a seeming inconsistency in the way people are answering the survey—in effect, they are saying that they personally have suffered less crime, but that crime overall is up. That can’t be true for the country as a whole, but it can be true for the Crime Survey. Leading candidates:
The Crime Survey deliberately ignores certain categories of crime, such as drug offences. These are (1) very common, (2) don’t have a direct victim, (3) have been de-prioritised by the police. So if I see increasing drug dealing around my area, I can say there is increasing crime, yet at the same time it won’t show up on how much crime I’ve suffered.
Biased sampling.
Respondents are measuring crime changes over a different (presumably longer) period than mere comparison of the Survey statistics would suggest.
On the other hand, perhaps there genuinely is a gap in perception. But it’s far from clear that this should be chalked up as people being “too pessimistic.” On the contrary, you are implicitly conceding that their model of their immediate community is accurate.
This, however, is very insightful.
Thanks—I should have checked the link more carefully. The links were more for illustration anyway.
Lots of people have claimed that we have a pessimism bias, and my post obviously assumes they are right, but it would be nice to see some proper empirical work on the issue.
Thanks!
I personally suffer approximately zero crime. I was pickpocketed (I think) about a year ago, and nonviolently assaulted the other day, and I don’t remember any other instances in two years in London.
If the crime rate is falling, I’m not going to notice that as an effect on me personally. I’m also not likely to notice the effect on me personally if it rises. My opinions about crime overall aren’t going to be strongly correlated with my own experiences of it.
I would throw in the third possible cause for the pessimism bias, besides the Argument From Crisis and the slanted news coverage. Many organizations (both governmental and private) have incentives to make the population more fearful. Few (and in case of governments, very few if any) organizations want to make the population fearless.
Examples are trivial: the police want the population to be afraid of crime because it gives them more power, guarantees their jobs, raises their status, makes compliance easier and oversight harder, etc. etc.
This principle applies widely and includes the top levels of the government—a scared population is just more malleable and more convenient to govern.
Does this come close to being a ‘group selection’ argument?
(A brief explanation just in case that phrase isn’t common knowledge: natural selection in biology works on the level of individuals, and individual genes. For ages, a minority of biologists have been trying to show that selection can also work between groups directly even if the adaptation harms the fitness of individual members of the group. So far, they haven’t been able to convincingly show that this exists.)
The police, as an organization, cannot ‘want’ anything, since police departments are not agents capable of wanting. So the question is whether individuals within the department share those incentives to a sufficiently strong degree that they are likely to justify and coordinate around goals like “I will spread a culture of fear throughout my neighborhood.” It might be that this is the case, but it’s less obvious at the individual level than it is at the level of the organization; I could just as easily see police enlistment as selecting for people who specifically desire to help people be less afraid.
They can as members of an institution because of the nature of people that engage in that kind of work. The tend to be 1s and 8s on the enneagram and they are into what is right, just, and correct. They drive around and they see ways they could solve problems and make people conform to their view of the world. Police want to be able to channel their anger at society and at crime at certain places. They want to have the tools they need to do that. The last thing they want is to do is see a criminal go free because they lacked a certain tool to arrest them or get information to arrest them. Creating fear around crime and the threat of pain and loss is how, institutionally, they can get what they want to do their jobs the way they think it should be done rather than the way that we the people might be able to consent to and approve. In this country, we have created fear around drugs, people of color, and given the police carte blanche.
I am not talking about natural or unnatural selection. I’m talking about incentives.
I don’t think this is a useful tack to take. By the same reasoning, organizations cannot ‘do’ anything as well. Reductionism is not a universal tool to be applied to everything you see.
They don’t need to justify and coordinate—they need just to do it.
You know you can test your seeing against empirical reality, right? :-)
Incentives work on individuals. If the individuals inside an organisation have no incentive but the group as a whole doesn’t, nothing gets done.
Funny how all the organizations are arranged in such a way that the organization’s incentives are transmitted to the individuals inside them...
Part of Alan Greenspan’s apology was that his biggest mistake was to believe that’s true. Organisations are structured in a way to incentivise behavior that benefits them but nobody gets a promotion as a police officer because his superiors think he advanced the interest of the police by getting the population to be more afraid.
You are mistaken. It is true that organization’s incentives are always transmitted to the individuals within it. What is not true is that those are the only incentives which these individuals have.
I wish it didn’t.
Turning from radicals to reactionaries, loss aversion could also support the perversity & jeopardy tropes Albert Hirschman identified in The Rhetoric of Reaction. Maybe loss aversion boxes us in from two directions: while it drives some people to conclude that things are worse than ever, it might discourage other people from trying to improve things.
Good! Actually Kahmenan argues that loss aversion and/or status quo bias is a mighty conservationary force (in Thinking Fast and Slow). I think that is quite right—generally loss aversion should make us fearful of trying out new things. Also progress is seldom pareto optimal—even thought the majority might gain, a small minority might lose. Due to loss aversion, they will be much more vocal than those who gain—which might stop progress in its tracks (cf. various Luddite protests against new technologies).
In fact, because of loss aversion, it’s hard to persuade people we need to change, in case you can’t argue that things have already changed for the worse—and we need to change them back!
Interesting point, hadn’t thought of that.
I’d like to put this in relation to the recently featured Cost for Irrationality post. I think this bias is a prime example of a bias capable of cripling a society.
I totally agree that this Argument from Crisis is (ab)used a lot. And I’m inclined to say that its massive use (obviously because it is so effective, even more so with our mass-media) has a profund effect on a lot of areas. I think it causes e.g. overprotection of children—making them more dependent and less autonomous with each generation.
I think the Argument from Crisis effect is able to persist across multiple generations precisely because it is not founded in facts but colored perception which may well be applicable over time. In fact it may even worsen because a perceived effect appears to continue. Except if it is explicitly countered, for which I see little general evidence though there may be specific examples. In the end you probably can’t argue for all time that some evil continues to increase when in fact it by now has almost been eradicated. Or can you?
For those who take fearmongering as a calling, moving the goalposts knows no limit.
The word you are looking for is nostalgia. The argument from crisis is very similar to restorative nostalgia ( http://samirchopra.com/2013/06/20/kinds-of-nostalgia/ ). The first study I found didn’t find much of a link between nostalgia and pessimism ( http://www.amsciepub.com/doi/abs/10.2466/pms.1998.87.2.411?journalCode=pms ).
I’ve seen Moldbug post of graph of assaults/stabbings etc… in England from 1900 to 2000. It went up by a factor of 30. You also have to consider that modern tech allows for the consequences of violence to be ameliorated. It’s genuinely hard to compare today to the past because we don’t live there and we have to rely on others accounts of it.
Could you provide a link?
http://www.significancemagazine.org/details/webexclusive/2518651/Crimes-of-the-century-Recorded-crime.html
It was like this.
It was there though the original pdf link is broken now.
Yeah, sure, because in the past life was so swell even for the bottom 75%.
Looking at the bottom line of this post I can say that one of the things are missing is who does the pessimism bias serve? It serves interests and those interests are interested in keeping a pessimistic view of the world going because they can make sales and increase their bottom line. They can sell commodities and make the public believe that they need to do things for themselves to make life better because things are bad. This noble myth keeps people in jobs they hate and keeps them docile. If the world was a bright and wonderful improving place then people would do radical things that would upset the social order and the accepted way of things. That is why radicals want radical change and why, packaged in the right way, they can become a representative of the whole. If you look at the wave of communism in the early 20th century, you had people getting behind the ideas of just a few radicals because of an overwhelming problem and a social order that was unwilling to change. That is what we face now. Facts have no place in this space because it is all about perception and perception is crafted. What you are describing is how people keep the status quo going.