behavioral economics – Hamilton Carvalho

What do fast food chains and behavior change have in common?

Consider the following questions.

In trying to change the behavior of vulnerable segments of our societies, why not use the same approach that fuels the successful global expansion of fast food chains (contributing to the current obesity and diabetes epidemics)? Why should the “devil” have all the best tunes?

In a nutshell, social marketing is a discipline that integrates knowledge from marketing and related disciplines on human behavior (like psychology, economics and anthropology) to inspire the creation of social programs that deliver value and change the behavior of individuals and segments of societies, increasing societal well-being as a result.

Social marketing enrich social programs by adding a unique value proposition: the ability to conjugate the understanding of human behavior with the use of the same techniques, principles and knowledge that companies employ to succeed in the marketplace (such as segmentation, branding and consumer research).

Social marketing builds a deep knowledge of social program’s targets from bottom-up, from understanding their values, beliefs, attitudes, and the barriers that prevents them from performing the intended behavior.

Since it was christened in 1971 as a legitimate offspring of mainstream marketing, social marketing has been employed with success to deal with varying social problems, such as improving health, protecting the environment, fighting diseases like HIV, decreasing the use of tobacco, and reducing poverty all over the world. Nowadays, it is a mature discipline and integrates the repertoire of several governments and other important social actors.

Maybe this talk reminds you of behavioral economics.

Behavioral economics has gained a deserved reputation over the last years, due to the Nobel Prize conceded to Daniel Kahneman and the late Amos Tversky and due to excellent books from academic stars such as Dan Ariely and Richard Taller (“Nudge”). It has shattered the concept of homo economicus, the idea that human beings are rational and powerful processing machines with predictable preferences. We must use their evidence-based insights on interventions to tackle complex social problems. But we should go further and use their findings under a more encompassing approach. That approach, that umbrella, with an incredible track of success in overcoming social problems is social marketing.

Social marketing also offers a kind of “portable” framework that can be employed to target social actors in charge of structural roles. In other words, most social problems have causes whose roots are distant from the individual level of behavior. Social change happens as a result of the interplay of several factors, many of them depending on the behavior of actors such as politicians, media professionals, pundits and the like. The flexible social marketing framework can help the mission of targeting them to create change at a broader and enduring level.

Did you know that?

Social marketing and complexity

This article briefly discusses some points I consider important for the future of social marketing. Let’s start with three popular behavioral models and then move to the question of complexity.

Health Belief Model, Transtheoretical Model and Theory of Planned Behavior are still common place, no matter several important methodological concerns about the first two and the inadequacy of the latter to account for behaviors highly dependent on situational or “irrational” forces. According to a 2008 paper, no research on HBM efficacy was conducted after one done in… 1984. TTM has constructs (like “dramatic relief”) with low validity and the process by which people transit through its stages is far from clear in its formulation. Segmenting targets according to their behavioral propensity is helpful, but TTM does not do a good job in explaining the processes behind the changes. A much better approach to behavior initiation, change, maintenance and abandonment is this one, inspired by the self-regulation stream of research:

Rothman, A.J., Baldwin, A.S., Hertel, A.W. (2011), “Self-regulation and behavior change: disentangling behavioral initiation and behavioral maintenance”, in Vohs, K.D., Baumeister, R.F. (Eds),Handbook of Self-regulation: Research, Theory, and Applications, The Guilford Press, New York, NY, pp.106-122.

I deeply recommend that chapter.

The permanence of simple models of human behavior in our toolbox speaks loudly about how ideas stick. The problem with them, however, is not only that they are rough theoretical approximations to social behaviors – approximations that were surpassed by theoretical developments in recent decades. The main problem, in my opinion, is that their simplicity belongs to an old paradigm.

Increasingly, complexity has been recognized as the main feature of the modern world, no matter to where one looks. In the last Peter Drucker Global Forum (in 2013) it was explicitly recognized that the management science does not have models to deal with the challenges faced by organizations. It was suggested that management theorists should learn from sciences more used to deal with complex phenomena, like physics. A recent survey by IBM with over 1.000 CEOs all over the world revealed that their main complexity challenge is… understanding consumer behavior!

According to a recent post by researcher Helga Nowotny on Harvard Business Review blog,

“what produces complexity is not so much the presence of many direct cause-effect links which operate with subtlety versus precision, but rather the presence of indirect, non-linear relationships between the variables, parts, and dimensions of the whole. What make complex systems so complex, therefore, are their multiple feedback loops and their indirect cause-effect relations which, moreover, play out at different speeds and on different time scales.” (http://blogs.hbr.org/2013/10/the-embarrassment-of-complexity/)

Complexity seems to occupy the center of attention in current management discussions (wicked problems, as Craig Lefebvre reminds us, are what we face in our activity). Complexity is everywhere. I was surprised the other day while reading the 2014 Harvard Health Report on Prostate (a well-written guide for patients). The subject of the introduction was… complexity! Wherever one looks one sees how complex phenomena dominate the social agenda. It has been recognized, for instance, that economic development depends on the existence of several concomitant conditions, that feed on each other (in complex interactions), like the development of institutions, societal trust, increases in productivity and educational levels, the quality of national strategic plans and their execution, the interplay of social and political forces, the balance of natural resources and so on. Not surprisingly, it is easier to reach a given standard of income (a challenge per se) than to overcome the so-called middle income trap, where several countries lie and from where they probably will never get out (I cite my own country and countries like China). Another example is corruption in a social system (organizations, societies), a social illness that propagates easily and it is hard to break, especially in developing countries. I used to think that corrupted people were evil people (a dispositional mindset) – committing the fundamental attribution error. Now, knowing the work of Philip Zimbardo and some behavioral ethicists (Max Bazerman, Francesca Gino and others), I see the interplay of several situational and cultural factors, at different levels, accounting for that problem.

Well, social behavior is a typical example of a complex phenomenon. Sometimes, public policy can rely on solutions that are paradoxically simple (ex. defaults in a choice architecture solution). While the solutions may be simple, the problems they intend to attack are complex and require a complexity approach. The key seems to be to embrace that complexity with a scientific mindset (requiring both an experimental approach to problems – what behavioral economists have done successfully – and a proper integration of the astonishingly immense – and disperse – knowledge on human behavior that has been accumulating over the last decades). To understand how people actually work, what makes them tick and how they change their behaviors it is important to integrate knowledge from a myriad of bodies of knowledge; ranging from the unsconscious, neuro research, evolutionary psychology, emotions, self-regulation, cognitive psychology, decision-making research, social psychology (including the power of situational factors), ethnography and so on… while, at the same time, avoiding the trap of ignoring upstream factors) To my knowledge, frameworks like EAST and MINDSPACE are an interesting but imperfect solution to this need of integration (I do miss on them a proper treatment of justice concerns and emotions – beyond the so-called affect heuristic). But these kinds of framework are the best we can hope for.

What is the takeaway? Reality (social, political and economic factors that comprise upstream challenges; human behavior) is complex and gray. Marketing and social marketing have a two-way relationship with consumer (human) behavior and depend on a proper understanding of that behavior – grasping what is a deeply counter-intuitive knowledge uncovered in recent decades.

To deal with complexity (…), says Helga Nowotny in the same piece above mentioned, “requires the ability to combine parts of the whole, however crudely, into an approximation of the look at the whole which we will never see entirely. It requires us to draw on the faculty of human judgment to focus on the smaller picture in order to comprehend the larger one.” She claims that we have to get used to an integrative thinking mindset, looking for patterns in a scattered set of data. I couldn’t agree more.

In another piece at HBR blog on the same subject, economist Roger Martin posited what I think it is the key paradox in understanding human behavior (http://blogs.hbr.org/2013/09/our-self-inflicted-complexity/): while the social problems are wicked problems with open solutions, almost all scientific knowledge on human behavior is compartmentalized, fostering inter-domain complexity, complicating the process of integration. This is the world where we live and work, where simple models of social behavior don’t cut anymore. But as long as we strive to integrate that knowledge, while acknowledging its complexity, I think we will be able to successfully compete within the behavior change arena. We will have to make do with imperfect solutions.

Confirmation bias and witch trials

We look to reality to confirm our beliefs. It is automatic, you cannot control it. We navigate the world looking for patterns that match our expectations and beliefs. Worse: We treat the disconfirming information in a way similar to how witches were “judged” during the incredible witch-craze that took place a few centuries ago (and represents one of several stains in mankind’s history.) Women accused of being witches often were subjected to what was known as the “cold water test”: they were thrown into a lake, for instance. If they floated, this was the “proof” they were witches and death by burning or hanging was their destiny. If somehow they did not float, they were considered innocent, but often they drowned and died anyway*. So a similar process occurs when we face some information that we don’t like or that disconfirms our mental models: We make sure (automatically) it will face a biased evaluation that eventually will kill it. There is plenty of scientific evidence showing this process in action. This is one of the human paradoxes we have to live with.

* Trevor-Roper, H. R. (1968). The European Witch-Craze of the Sixteenth and Seventeenth Centuries and Other Essays.

O desconto na conta de água é insuficiente!

A mente humana não consegue conviver bem com a incerteza e exige respostas definitivas. Estamos vivendo já as piores consequências do aquecimento global? A ciência ainda não consegue dar respostas com certeza absoluta a questões como essa, embora já exista um consenso entre a maioria dos cientistas dedicados ao tema de que o aquecimento global é fato. “Mas como?” – Grita o cidadão que recebeu um e-mail ou assistiu um vídeo no YouTube mostrando o ponto de vista de alguns cientistas céticos. “Esse aquecimento global é história para boi dormir”, diz ele. Reações como essa, que são comuns, indicam apenas o desconhecimento do método científico. A unanimidade é inimiga da ciência, sem dúvida. Mas uma abordagem racional dos problemas sociais demanda que se aceite o consenso da maioria dos cientistas, gerado a partir do conhecimento produzido em pesquisas que foram revisadas e publicadas nos principais periódicos acadêmicos da área. É bom que se repita: A maioria absoluta dos cientistas dedicados ao tema está suficientemente convencida tanto do aquecimento global quanto de seus potenciais efeitos negativos para a vida no planeta. Como explica o estatístico americano Nate Silver, no excelente The Signal and The Noise, a extrema complexidade inerente ao estudo de campos do conhecimento como a climatologia torna virtualmente impossível fugir de abordagens como simulações rodadas em computador (que podem ser muito poderosas) e a construção de cenários. Em um desses cenários, estruturado por um grupo de cientistas do MIT, existe uma probabilidade razoável de consequências assustadoras para a vida na Terra, reflexo de uma elevação da temperatura média de 5 graus Celsius até o final deste século.

Mas o que isso tudo tem a ver com a decisão da Sabesp de conceder desconto na conta de água para os consumidores abastecidos pelo Sistema Cantareira? O histórico verão com clima de deserto em São Paulo secou os reservatórios e pegou, ao que parece, todos de surpresa. Fala-se sobre a possibilidade de racionamento, o que as autoridades negam – um claro indicador de que a possibilidade é real. A resposta da Sabesp para tentar diminuir a demanda foi dar o desconto de 30% para quem economizar ao menos 20% em seu consumo. Mas essa medida é claramente insuficiente para produzir uma redução significativa do consumo. O principal problema dela é estar calcada em uma certeza absoluta na mente dos gestores públicos, mas que vem sendo demolida já há algumas décadas pela economia comportamental: A ideia do homo economicus, cuja mudança de comportamento na direção desejada depende apenas da existência de incentivos econômicos adequados. Sim, não há dúvidas de que as pessoas respondem a incentivos, como regra geral. Mas, além dos casos em que os incentivos funcionam na direção oposta da pretendida (o que não vou abordar aqui para não tornar o texto muito longo), a ciência comportamental vem comprovando ao longo das últimas décadas que as pessoas respondem também a outros fatores importantes, frequentemente de forma mais poderosa do que a incentivos financeiros.

A situação que a Sabesp enfrenta, convenhamos, não é fácil. A água sempre foi encarada pela sociedade como um recurso infinito. O desperdício é comum e o preço da água não reflete seu potencial de escassez. O problema também se agrava pela crescente verticalização das cidades, que resulta em prédios cuja medição de consumo geralmente não é individualizada, gerando um poderoso incentivo ao consumo elevado. Em um cenário de racionamento de água será comum o comportamento de carona (free rider) ou o chamado efeito Ringelmann – imagine que você está puxando a corda de um cabo de força e que há mais 8 pessoas com você – o esforço individual acaba sendo menor. Poucos economizarão de verdade, enquanto os demais condôminos manterão seu padrão de consumo.

O que outros países que enfrentaram situações de escassez de recursos fizeram? As melhores experiências mostram, por exemplo, que fornecer um inequívoco feedback comparativo do padrão de consumo de uma residência ajuda a diminuir o consumo. O feedback precisa ser concreto, rápido e precisa refletir uma comparação com residências similares ou com vizinhos de um mesmo bairro. Nos Estados Unidos, reduções significativas de consumo de energia elétrica foram obtidas com o envio de contas de luz com a comparação do consumo com a média da vizinhança, bem como com a residência que mais reduziu o consumo, além de informações práticas para alcançar o objetivo pretendido. A utilização de “carinhas” (feliz, neutra, triste) que indicam o nível de consumo em relação ao desejável também tem sido bem sucedida, bem como a estruturação de comunidades virtuais com níveis de desafio (no caso, a economia de água) progressivos, que podem ser galgados pelos consumidores na medida em que determinados comportamentos são adotados – a recompensa é a atribuição de mecanismos simbólicos de reconhecimento (badges). Outras iniciativas internacionais incluem, por exemplo, intervenções para criar jardins e campos que demandem pouca água (com a escolha de plantas e mudanças no sistema de irrigação), incentivos para a troca de vasos sanitários, torneiras e chuveiros, programas educacionais em escolas e em comunidades (com foco prático), adoção de selos de economia em equipamentos, atividades direcionadas à faixa de residências com maior consumo (como o emprego de compromissos escritos) e a divulgação da experiência de pessoas bem-sucedidas (role models) na adoção do comportamento desejado. O foco dessas abordagens é sempre a oferta de benefícios tangíveis e intangíveis ao consumidor, de modo que este escolha por si só aderir aos comportamentos esperados. Existe um consenso na literatura internacional dedicada ao tema: É preciso uma abordagem sistêmica, baseada em abordagens múltiplas, simultâneas e alimentadas pelo conhecimento científico sobre os motores do comportamento social.

Além dos exemplos listados acima e que poderiam ser adotados no esforço de redução de consumo, seria recomendável à Sabesp o investimento no desenvolvimento de soluções acessíveis de medição individualizada de consumo em condomínios. As soluções existentes hoje parecem ser caras, dependendo do tamanho e da estrutura do condomínio. A parceria com empresas que atuam nessa área e com centros de pesquisa é um caminho interessante.

É possível que no futuro a água potável seja tratada como um bem tão escasso como o petróleo e que sofra um choque de preços para alterar a demanda? Não tenho resposta para isso, mas tenho a convicção de que é preciso estar preparado para um cenário desse tipo. Precisaremos de uma mudança significativa nos padrões de consumo da água, que pode ser alcançada com a ajuda de medidas simples, como demonstra a experiência internacional. Só o apelo ao consumo racional, como tem feito com frequência o governador de São Paulo, ou o desconto temporário na conta são claramente insuficientes para mudar o comportamento dos consumidores.

Social marketing versus behavioral economics

I see a great deal of overlap between social marketing and behavioral economics and, of course, some differences.

Both disciplines rely on making things simpler. As I use to say, if you want to make a behavior popular you need to make it screamingly, conspicuouslysimple and easy. This focus fits usual interventions inspired by both approaches.

Another similarity is applying the same mindset notwithstanding differences in the tool name. In most cases I cannot see how choice/social architecture differs in nature from deploying some of the social marketing techniques (removing barriers, place strategies) to facilitate the performance of the target behavior. For instance, a program like Save More Tomorrow could be conceived under either conceptual umbrella.

In “The Political Brain”, Drew Westen stresses the role of associations and narratives (woven through emotional processes) in deciding the fate of political candidates. Of course marketers see easily the importance of associations and narratives for any brands or attitudinal objects. The way I see it,behavioral economics has achieved prominence mostly propelled by a strong narrative (and related metaphor): the David that overcame the Goliath (traditional economics – or better – rational homo economicus) destroying in the process one more facet of the idealistic vision of human beings (a trend centuries old). This narrative was strengthened by the Nobel Prize won by Daniel Kahneman in 2002 and the initial contempt (still existent) the discipline faced in traditional academics departments. Richard Thaler’s work with UK government and the building of a practical framework (Test, Learn, Adapt is their motto) to extend scientific method and mindset to pressing public issues has added to their growing popularity. The more common intuition is contradicted by the results of their experiments the more behavioral economics demonstrates the advantage of the scientific method (and increases its fame).

Behavioral ecomomics has discovered some fundamental truths about human decision making and behavior (beginning with Kahneman and Tversky’s Prospect Theory and its offsprings: framming effects, loss aversion, mental accounting, hyperbolic discounting and so on). In turn, the body of knowledge and the prestige the discipline gathered over time gave birth to specific ideas to tackle important public issues and well-succeeded interventions. But behavioral economists have been drawing heavily on other streams of research on human behavior, like (broadly speaking) social psychology. They seem to be successul in integrating theories and findings in areas like emotions, mapping, dual-systems, social norms, choice overload, implemmentation intentions, fairness, construal and the identifiable victim effect. I think a great deal of their deserved reputation comes from this integration. We are all aware of Cialdini’s (and related researchers) work on social norms. Dual-systems approach (System 1/System 2) has been researched by scholars under the psychology/philosophy of mind tradition, like Keith Stanovich (University of Toronto) and colleagues. And so on. But few practical approaches have been developed besides behavioral economics and social marketing.

I see social marketing as the most powerful social technology to address the social problems faced by our societies. My academic advisor used social marketing to structure important governamental programs, dating back to 1981. A bit unlike behavioral economics, it has a parsimonious and flexible toolbox that can easily acommodate and integrate the different theoretical avenues that offer insights on human behavior. Overall, social marketing is a more encompassing approach. From my point of view, however, I feel social marketing needs a more compelling narrative. Last month I read a recent book on tax-related behavior. It had several mentions to behavioral economics. It did mention segmenting taxpayers (something developed countries have been doing for more than a decade) – but there was no mention to marketing as inspiring that approach. I feel social marketing needs a better positioning.

The challenge is enormous and daunting, but I see it as feasible as the discipline matures.

There are important streams of research calling for some kind of integration – even if imperfect. I cite Self-Determination Theory (and one recent practical offspring: gamification), research on Self-Regulation and self-control (with promising understanding of behavior evolution and maintenance – a great advance over transtheoretical model – I mentioned it briefly in my recent paper in JSM), positive psychology, applied neurology… the list of course is not comprehensive.

I see the challenge as threefold: develop a compelling narrative, integrate findings from promising streams of research and dialogue with behavioral economists and other important scholars – given their prominence and given what seems to be lack of knowledge about social marketing. This open dialogue and the transdisciplinary approach can enrich social marketing, in my opinion.

Eu não me lembro o que você fez há 10 meses: vieses e avaliação de desempenho

Um livro que indiquei em outro post (Decisive: How To Make Better Choices in Life and Work) apresenta soluções muito práticas para superar diversos vieses que atingem o cotidiano das organizações, frequentemente com consequências sérias.

Um outro livro que trata do mesmo tema (e que compila como poucos os vieses identificados por décadas de pesquisa na economia comportamental e na psicologia social), também com enfoque prático, é o livro dos professores de Administração Max Bazerman (Univ. Harvard) e Don Moore (Univ. Carnegie Mellon). Para quem tem interesse no tema, é leitura imprescindível: Judgment in Managerial Decision Making (2009). É utilizado no curso que o Bazerman dá em Harvard.

O livro de Bazerman e Moore trata, por exemplo, de um viés que tem consequências na avaliação de desempenho, realizado anualmente nas organizações. Como esse viés se manifesta? Com base em sua memória, gestores avaliam o desempenho de funcionários de forma naturalmente enviesada: exemplos mais vívidos de comportamentos dos funcionários (sejam positivos ou negativos) são mais facilmente lembrados e vão dar a impressão de que são muito mais numerosos do que os comportamentos comuns do dia-a-dia, que não são gravados na memória do gestor da mesma forma. Assim, os comportamentos mais facilmente recuperáveis da memória e que tendem a não ser representativos do desempenho do funcionário terão um peso claramente desproporcional em avaliações de desempenho. A proximidade temporal dos comportamentos também tem uma influência comprovadamente desproporcional: as evidências mostram que os gestores dão um peso muito maior aos eventos ocorridos nos últimos 3 meses do que nos 9 meses anteriores. Destaco: mesmo assim, a maioria esmagadora das empresas faz apenas avaliações anuais. Esse é mais um caso de gap entre o que a ciência sabe e o que as organizações aplicam.