Our Work

Behavioural Insights and Crime: Part I

Violence against Women (VAW)

When you think of BI and crime, shows like Criminal Minds and Mindhunter (both of which the author of this blog post highly recommends) come to mind. However, tackling VAW requires more than psycho-analysing perpetrators. This is because there are a multitude of other factors leading to the creation of a climate of acceptability toward VAW. Here are some examples, drawing largely on a report by the European Commission:

  • Victims of violence may not report the crime: They may be afraid of the perpetrator, believe they are unlikely to receive help or simply because they are not aware of the support services available. It is also possible that a status quo bias is at play here; victims may be inclined to accept the situation and be unwilling to change it. Social norms could also prohibit them for speaking up.
  • Bystanders may not intervene: Either because they believe it is not their responsibility or that violence within couples is a private matter. They may also lack self-efficacy, i.e. confidence in their ability to make a difference.
  • Perpetrators may not believe they committed a crime, or think they can get away with it. They may also be present biased about the effects on the victim as well as implications for their own lives (punishment/ imprisonment)
  • Disagreement on definitions: For example, while most people agree that rape is wrong, there may not be a consensus on what exactly constitutes rape.
  • Professionals within the legal system may have prejudices too, creating an environment that is not supportive for victims to come forward.
  • Stereotyped media portrayal or inappropriate reporting by journalists of news regarding VAW.

rape

Source: emaze.com

The list is endless and usually a combination of factors operates to encourage victim blaming attitudes. Which is why in the mid-1990s, the causes of VAW were recognised to be probabilistic rather than deterministic. That is to say there is no single cause; the same outcome (VAW) can be caused by a different amalgamation of factors in different social contexts.

One of the simplest models for analysing behavioural causes for a certain action is the Theory of Planned Behaviour. It states that for someone to perform a given behaviour, the following 3 conditions have to hold (assuming they have the intention to carry it out):

  1. Holding a positive attitude toward the behaviour
  2. Considering the behaviour to be in line with social norms
  3. Having self-efficacy i.e. believing that they are able to perform the behaviour

How does BI fit in?

The aforementioned examples can be viewed in light of the above model to design programmes for changing attitudes and behaviours. Some behavioural levers that can be used for this purpose are:

  • Using social norms:

Example: In 1999-2000, James Madison University (JMU, United States of America) ran a campaign aimed at changing misconceptions among male college student about their peers’ sexist beliefs. The campaign used a series of posters and flyers containing contextualised normative messages like ‘A man always prevents manipulation: three out of four JMU men think it is NOT okay to pressure their date to drink alcohol in order to increase the chances of getting their date to have sex’ or ‘A man respects a woman: nine out of ten JMU men stop the first time their date says “no” to sexual activity’. Results showed that there was a significant increase in the percentage of males claiming that they ‘stop their sexual activity as soon as their date says no’, and who endorsed the statement ‘when I want to touch someone sexually, I try and see how they react’.

  • Scarcity:

Victims of violence are in a situation of scarcity of mental resources: they can be under severe time, emotional and/or financial constraints, impairing their decision-making abilities. Therefore making a specific plan with concrete steps to be carried out and creating awareness about it can enhance their self-efficacy.

  • Framing:

The narrative around VAW is slowly changing, with several countries changing the legal definition of rape from “crime against morality” to “crime against the individual”. Similarly, interventions can be designed to convey that VAW is not a private matter and is a serious crime.

Example: A 2002 advertising campaign by New York City focused on increasing reporting by women experiencing domestic violence via a 24-hour telephone hotline. Behavioural levers included framing messages to highlight that violence is a crime for which there is no excuse, and that abusers are diverse and include men with a positive image in society. Posters showed pictures of men — typically a college athlete or professional businessman — behind prison bars, with headings such as ‘Employee of the month. Soccer coach. Wife beater’ or ‘Big man on campus. Star athlete. Abusive boyfriend’, along with the subtext ‘There’s no excuse for violence against women. Men who hit or abuse their partner belong in jail. Report domestic violence and get the help you need’.  Results showed that calls to the hotline increased by 36% in the second week of the campaign.

  • Loss aversion:

Perhaps policemen and judges could be educated on the cost of VAW, rather than focusing solely on reducing prejudice and emphasising the importance of women’s human rights. Also, potential assailants could be discouraged by highlighting everything they stand to lose- their freedom (if imprisoned), their reputation (although this would mean shifting social norms to shaming the perpetrator rather than the victim) and so on. Victims could also be encouraged to report the crimes against them by showing them how they would be losing out the chance to live an abuse-free life.

There are several other behavioural levers that can be harnessed, like choosing the source of the message wisely, addressing the status quo bias, etc. Another idea for changing attitudes, inspired by Dan Ariely’s Ask Ariely column could be presenting people who hold misogynistic views with even more extreme arguments supporting their belief. The absurdity of the ideas may force them to re-evaluate their own attitudes.

Conclusion:

We are finally realising the importance of behavioural science in designing interventions for reducing crime, including VAW.  It is important to pretest initiatives before implementing them and also conduct an assessment of impact once it is in place. VAW is a complex phenomenon, but if we break it down and target specific behaviours, perhaps attaining the goals we want to is not as unattainable as it may seem!

 

Advertisements
Article Review

“Adapting to the Destitute Situations: Poverty Cues Lead to Short-Term Choice”

Review of a paper by Liu, Feng, Suo, Lee, Li (2012)

In previous posts (“Scarcity”, “Through the psychology of poverty”), we showed how scarcity affects the way people consider problems and take decisions. Although scarcity may concern many kinds of goods, including even available time, in this article we will consider only the lack of financial means. The underlying idea is that when people are wealthy enough not to have to worry about small expenditures, they do not need to spend mental resources on the task, whereas indigents have to invest time and thought on it.

This has a consequence: scarcity leads people to focus primarily on problems wherein shortage is more severely perceived, even though those same issues are not necessarily the most relevant ones for long term well-being. This compromises the individual’s ability to choose wisely.

Mani, Mullainathan, Shafir and Zhao (2013), following the same intuition, showed how Intelligence Quotient (IQ) test scores are correlated to the perception of poverty. In particular, by inducing thoughts about everyday small financial problems in poor subjects, they were able to severely diminish their cognitive ability, while leaving the wealthy ones unaffected. Interestingly, when no manipulation was performed, there were no significant differences between the rich and the poor’s test scores, suggesting that the reduction in cognitive skills was due to the stress linked to the fact of being (relatively) poor, and not to indigence itself. The idea behind this observation is that cognitive capability is limited and poverty – or simply a reminder of the fact of being poor – “taxes” it, leaving less to handle everyday choices.

Liu, Feng, Suo, Lee and Li (2012) took the discussion a step further: in this case, the considered subjects (all college students) were not poor, but simply cued into different economic statuses. In particular, they studied whether poverty cues affected inter-temporal choices under the general assumption that people exposed to poverty prefer a short-term but smaller reward.

In all experiments, participants could opt between a smaller but immediate payment and larger but later in time one. They were asked to take their choice before and after completing the priming task, so that the first one can be used as a benchmark. The games were repeated several times. In Experiment 1, subjects were primed in an explicit way by having to judge the degree of poverty or affluence (according to the group they were randomly assigned to) of several pictures; in Experiment 2 they had to count the number of people in each picture, so that the purpose of the experiment was less explicit; in the third one they participated in a lucky draw game in which they would gain a prize or nothing, thus being exposed to a moment of relative affluence or poverty.

22782124_10212901226357233_1588832224_nPictures representing “poverty” and “affluence” from Experiments 1 and 2.

In every experiment the groups assigned to the different cues were balanced, as the percentage of present biased individuals was the same. However, after completing the task, subjects who had been exposed to poverty cues became significantly more prone to opt for an immediate but smaller payment, while those cued into affluence showed a non-significant increase in the choice of the later but larger prize. Results were the same in each experiment, thus suggesting that the environment influences individuals’ perceptions, which is in turn reflected in time preference. Indeed, according to the authors, the poor are subconsciously associated with an unstable context and a lack of means to deal with it, whilst the rich are perceived as economically independent. As a consequence, people, when exposed to situations of poverty, felt more in need of liquidity to deal with uncertainties and hence opted for a smaller but closer-in-time reward. Moreover, the fact that the choices of the other group did not change suggests that individuals are more sensitive to negative cues.

22833343_10212901224477186_1592177517_o22812843_10212901225637215_307340708_oResults from Experiment 2 and 3 respectively: mean percentage of immediate reward as a function of the manipulations of the ‘‘poverty’’ state (left) and the ‘‘affluence’’ state (right), with pre-test (brown) vs. post-test (green) percentage of demand for immediate payments. Error bars indicate standard errors of the mean.

Therefore, the authors conclude that  “just the feeling of poverty influences intertemporal choices – the actual reality of poverty (restricted resources, etc.) is not necessary to get the effect”.

 

References:

Liu L., Feng T., Suo T., Lee K., Li H. (2012), “Adapting to the Destitute Situations: Poverty Cues Lead to Short-Term Choice”, PLoS ONE, Vol. 7, No. 4: e33950.

Mani A., Mullainathan S., Shafir E., Zhao J. (2013). “Poverty Impedes Cognitive Function”, Science, Vol. 341, No. 6149: 976-980.

Mullainathan S., Shafir E. (2013), Scarcity: Why Having Little Means So Much, Times Books, New York.

 

Our Work

When our brain decides for us… And without our permission

(Originally published in Slovak at mindworx.net)

Most of us probably believe that we are in control of our own decisions. We have our opinions, beliefs and principles, we know what we like and dislike and we always decide in accordance with our preferences.

However, this is not entirely true. Our brain reacts to all kinds of cues from the external environment and so our behaviour and decision making depend heavily on these. Even if we do not realise it.

MINDSPACE FRAMEWORK

Since there are a lot of these external cues, it is useful to categorise them to be able to analyse them better. One of the possible models is the “MINDSPACE” framework, created a few years ago in the UK by the Cabinet Office and the Institute for Government. Obviously, such models do not include all the possible factors that influence our behaviour, but do describe the most important ones. Thus, it serves as a good basis for a better understanding of human decision making.

The name MINDSPACE is an acronym made of the first letters of the main motivators of human behaviour. These are Messenger, Incentives, Norms, Defaults, Salience, Priming, Affect, Commitments and Ego. Let’s now have a look at each of them.

mindspace-1

Messenger

It is true that how we perceive information depends on their source. Different researchers have shown that we are more likely to listen to what the experts say, but also to accept advice from people similar to us in some ways. On the other hand, if we get advice from someone we dislike, it is probable we will not take it into account no matter how good it is.

Incentives

They include everything that motivates us – either physical motivators (such as money or other non-financial rewards) or psychological motivators (like our intrinsic motivation to do something good for the others). It is useful to know how we react to various motivators and which of them work best in various situations. In general, it is true that losses loom larger than gains, so a potential loss can be a better motivator than a potential gain. Also, our perception is relative – the same amount of money can be seen as too small or too large depending on the reference point.

Norms

More precisely, social norms. People tend to be influenced by what other people are doing. Usually nobody wants to be the “outsider” and have fingers pointing to them, just because they are different.

Defaults

People tend not to change the pre-set options, even for important decisions. The countries in which you are automatically considered to be an organ donor, unless you opt out, have a much larger proportion of donors than the countries in which you have to opt in to become one.

Salience

Logically, our behaviour is influenced by the things we pay attention to. And since there is a huge number of stimuli out there, our brain has to be selective – it pays attention mostly to what is new, simple and different. Sometimes, though, the brain actively seeks for cues that facilitate the decision making. For instance, it may search for an “anchor”, which is the information (mostly numerical) that is used as a basis for making a (numerical) judgement. However, such information does not need to be relevant in a given context and even a randomly picked number can influence how we decide.

Priming

Our behaviour is influenced also by our senses and at the unconscious level. The exposure to a certain type of words, sounds or smells can have a large impact on our behaviour. For example, the participants of an experiment that were required to read words related to old age left the room in a slower pace than the other participants. Or the smell of an all-purpose cleaner in a school canteen prompted the students to leave their tables cleaner.

Affect

It is nothing new that the emotions play an important role in our decision making. What is a bit more surprising is the fact that they influence us even when we do not realise it. An experiment has shown that when a mortgage offer included a picture of a smiling woman, the demand for the mortgage increased in the same amount as if the interest rate of the mortgage decreased by 25%.

Commitment

Commitments and promises, mainly the public ones, determine how we behave. If you struggle to do something, it is useful to commit yourself in some way. Tell your friends that you will stop smoking by a certain date. This will force you to do everything to accomplish our goal, just to avoid the shame of admitting the failure in front of people that are important to you.

Ego

Our actions should be in accordance with what image we have, or we would like to have (in our own eyes, or in the others’ eyes). For instance, men are more willing to contribute to charities if they are approached by attractive women, because in this way they manage to keep a good image in the eyes of the opposite sex.

SO, WHAT?

The MINDSPACE model has been created mainly for public institutions with the aim of improving public policies. Many governmental programmes are unsuccessful exactly because they are not designed for real people, but for “people” that are described by the theoretical (mainly economic) models.

This does not mean, though, that the framework is useless outside the public sector. Each one of us can learn a lot from it. If we better understand how our brains work and what influences our behaviour, we will be able to make better decisions, develop more successful products or provide better services and have better relationships with people around us.

Source: https://www.instituteforgovernment.org.uk/sites/default/files/publications/MINDSPACE.pdf

Article Review

“Scarcity”: a book review

What is scarcity?

“Scarcity” was published in 2013, written by Sendhil Mullainathan, professor of Economics at Harvard, and Eldar Shafir, professor of Psychology and Public Affairs at Princeton. The greatest strength of this book is also its greatest weakness: it’s educational. On the one hand this is good, because it brings Behavioral Economics to the general  public and we certainly need it; on the other hand, the lack of a model makes the analysis shakier than we would like.

At the very beginning the authors define scarcity as “having less than you feel you need”; traditionally, development economists focused only on physical poverty. However, with this definition, they want to emphasize that scarcity can be both physical and mental. The main takeaway of the book is that scarcity is a mindset that lures people into a poverty trap, the leaving of which is very difficult.

To better illustrate this point, throughout the book they refer to  several experiments that have taken place both in developed and developing countries. The best feature is that they test physical and mental scarcity in both settings, and more importantly how they interact. The interaction is fundamental, because their thesis is that physical scarcity causes mental scarcity, i.e. the scarcity mindset. It’s easy to imagine physical scarcity, it ranges from not having enough food, money, time or (spoiler alert!) enough blueberries in a videogame. However, it is more difficult to think of mental scarcity, even though we constantly experience it. To explain it, they use the concept of bandwidth, or mental capacity. We can think of it as the “RAM of our brain”, the precision and speed at which we process information and execute decisions.

What are the effects of scarcity?

When we experience  scarcity we tunnel, i.e. we get extremely focused on solving the scarcity at hand. This focus can have 2 effects: one positive, and one negative. In the short run it is generally positive, because it makes us focus on the task at hand, operating at maximum capacity. However, it can have a negative effect too, since in this case when we are focusing we are actually tunneling, i.e. leaving everything else out of the picture, with possibly dire consequences. Moreover, in the long run scarcity erodes the availability of mental capacity, given that we can focus on nothing else.

To support this view, they refer to  several experiments and I’m going to explain three of them to help you understand. There is one experiment where people with less bullets available in each round of a shooting  game scored proportionately higher than people with more bullets: having less prompts efficiency. In another one, dieters found more difficult to concentrate on the next task if the previous one mentioned pastries: scarcity captures the mind and lets us focus on nothing else. The final experiment concerns Indian farmers, who score lower on an IQ test (measured with Raven’s matrix test) just before harvest than just after it. Why? Because they typically squander their harvest money in the months immediately following it, and the closer they get to harvest period, the poorer they are. That causes what is called a “bandwidth tax”; part of the brain continuously focuses on the scarcity at hand, leaving less room for other thoughts. It is really important to notice that all kinds of scarcity could cause that tax, ranging from money, time, love, etc.

Before turning to how to escape the scarcity trap, they focus on another negative side effect: borrowing. When tunneling, we leave everything else outside the tunnel, such as future consequences. To illustrate it , they make their case against the utility of payday loans. These loans are extremely short ones, with high interest rates. Moreover, they are extremely easy to get: in the US there are more payday lenders than McDonald´s and Starbucks outlets combined! Hence, once inside the tunnel, payday loans seem like the best way to get money, but in the long run they’re extremely detrimental.

How can we help people escaping the scarcity trap?

Seeing as scarcity is a lack of something, there are essentially 2 solutions: a windfall of the something needed or a more efficient use of the quantity at our disposal.

However, scarcity is insidious. For instance, they once gave Indian traders enough money to extinguish their debts, drastically increasing their disposable income, since they did not need to pay interests anymore. However, one by one they started borrowing again, eventually reverting to their previous situation. So a one time payment wasn’t enough to escape the poverty trap because they were not efficient, i.e. they did not reserve any slack. Indeed, the main reason to fall into the scarcity trap is that we are at the limit of the resource in question, so that when we face an unexpected “expense” we start borrowing from the future and taxing our bandwidth. Hence, the authors strongly advocate for the creation of slack, both to exit and to avoid (re)entering the scarcity trap.

Moving on to the strictly efficient part, they refer of the St.John’s Regional Center in Missouri: they were  constantly operating at maximum capacity, so when an unexpected surgery came up all the elective ones needed to be rescheduled. The solution was brilliant: keep an Operating Room (OR) only for the unexpected surgeries. Having one less OR available actually increased the number of surgeries performed: the creation of slack solved the problem. The last example highlights another important feature of scarcity: it does not concern only individuals, but also organizations.

In conclusion, the authors offer an exciting, novel interpretation of  poverty, tackling it from different angles, following a new framework: scarcity. Even though they lack a formal model (so far), the book offers an interesting view on one of the oldest problems, using several hints from Behavioural Economics. In short: a must read.

Article Review

Through the psychology of poverty

What explains the differences in economic decisions amongst poor and rich individuals?

In 2014 Johannes Haushofer and Ernst Fehr, professors at Princeton and Zurich University, respectively, have written the article “On the Psychology of Poverty” for the Science magazine (original version here). They show evidence that poverty causes psychological consequences such as negative affectivity and stress with unexpected changes in economic behavior by changing individuals’ revealed preferences and leading to short-sighted and risk-averse decision making. But what are the channels through which poverty could arise and perpetuate itself?

Bear in mind two things when interpreting their findings. First, although poverty is defined as the lack of sufficient income, it is also characterized by the exposure to dysfunctional institutions, violence, crime, poor access to health care, etc. Second, being born in such an environment can trigger processes that reinforce poverty. For instance, lower willingness to take risks, adopt new technologies, invest in education and health, together with present-biased income preferences, make it harder to escape from poverty.

The effect of poverty on economic behavior. Some laboratory experiments have randomly assigned individuals to income shocks after they have earned some money in an effort assignment. Then, researchers compared the discounting of future payoffs between “treated” individuals with negative income shocks and the “control” group, who didn’t experience any changes. Such exogenous manipulation of income eliminates any potential reverse causality between discount rates and income. They found that individuals with negative shocks on income showed more present-bias behavior than others. No analogous effect was obtained for positive shock on income.

The effect of poverty on psychological characteristics. Recent findings show a positive correlation between income and happiness/life satisfaction both within and across countries (see fig.1).

1
Fig.1. Relationship between income and life satisfaction within countries

According to the World Health Report, the poorest population quintiles in rich countries have records of depression and anxiety disorder up to twice as large as that of the richest quintiles. Besides, there are numerous findings that poverty is positively correlated with the stress hormone cortisol, as well as depression, anxiety and unhappiness. One should already expect such relationship, but is it a causal one?

 

Causal effect of poverty on affectivity and stress. A study by Haushofer and Shapiro (2013) evaluated the effects of an unconditional cash transfer program in Kenya on psychological well-being, by measuring characteristics such as happiness, life satisfaction, depression and stress. Individuals were randomly selected to receive transfers of $0, $400 or $1,500, and they found positive effects for all variables whenever receiving any positive transfer. However, levels of the stress hormone cortisol decreased only for the ones receiving a large transfer.

 

2
Fig.2. Z-score happiness response and levels of stress hormone cortisol

A series of natural experiments (e.g. lottery payouts, introduction of guaranteed incomes, access to pensions) suggest causal links between increases of income and well-being (e.g. reduction in hospitalization, lower consumption of anxiolytics, increased self-reported mental health). Randomized control trials also show that offering households access to health insurance, better housing conditions and access to water have a significant effect on psychological well-being.

The takeaway from this review is that the poor may intrinsically have identical time and risk preferences to those of wealthier people, but their discount rates and risk-taking behavior can change if living in a chronic condition of poverty. Should a welfare state intervene to avoid the creation of a poverty trap? According to the authors, there are three possible courses of action to break this vicious circle, which requires directly targeting:

  • poverty through poverty alleviation programs (e.g. cash transfers), proven to increase general welfare;
  • its psychological consequences (e.g. interpersonal psychotherapy);
  • its deriving economic behaviors (e.g. commitment savings account, reminders to save), which usually lead to considerable increases in savings.

These interventions allow us to better understand the relationship between poverty, its consequences and their potentially negative effects on decision-making, giving us a fresh perspective on how development economics can help to tackle poverty.