Mukhopadhyay iems leanwashing 02 march2017

transcript

Lay Theories of Obesity: Corporate Leanwashing and the Obesity

Crisis as Market Failure

Anirban MukhopadhyayHKUST Institute for Emerging Market Studies (IEMS)

Academic Seminar Series

A scientific theory

Lay Theories

• “Implicit assumptions that ordinary people hold about themselves and their world” (Dweck 1996)

– Help give meaning to ones environment (Wyer 2004)

– Have systematic and predictable influences on behavior

• Everyone holds lay theories in numerous domains, e.g.,– Lay theories about the causes of failure – whether it is attributed to

intelligence or effort – predict helplessness, drops in self-esteem, and eventually, grades (Robins and Pals 2002)

Today’s focus: Lay theories of obesity

What, do you believe, is the main cause of obesity?

• Poor diet

…or…

insufficient exercise?

McFerran and Mukhopadhyay, Psych. Science 2013Karnani, McFerran and Mukhopadhyay, CMR 2014, JACR 2016

… and ongoing

This talk in one slide

• The more someone believes that insufficient exercise [poor diet] causes obesity, the heavier [lighter] they are actually likely to be.

– Controlling for a lot of things…– …Because exercise theorists eat more

• Some of this is attributable to Marketing– In an indirect way that had not been realized till recently

• There is an urgent need for decisive action

Marketers Love Food (and Vice Versa)

• The food industry is among the top advertisers in the U.S. media market (Story et al. 2002).

– Over $4.2 billion on fast food alone

• One third of the advertising in children’s TV programs (Desrochers and Holt 2007).

– 72% for candy, cereal, and fast food (Mellow et al. 2006).

• Very effective (Wilkie and Moore 2003; Dhar and Baylis 2011).

• “The message communicated … is that unhealthy eating (e.g., frequent snacking on calorie-dense and nutrient-poor food) is normal, fun, and socially rewarding.” (Chandon and Wansink 2011)

• Food quality is emerging as a major issue worldwide– Adulteration– Factory farming

• Safety issues: Monocultures• Nutritional issues: Loss of nutrient diversity• Environmental issues: Land usage• Ethical issues: CAFO (Concentrated Animal Feeding Operations)

– GMOs– Labeling

• Do people notice the labels? Do they read the labels? Do they understand the information? How do they use it?

– Additives

People Love Food (but Vice Versa?)

Portion Sizes = More of the Bad Stuff

Inflation at McDonald’s

1955 2000s

Ounces

McDonald's Soda Portion Sizes

Child's (110 cal)

Small (150 cal)

Medium (210 cal)

Large (310 cal)

2.02.5

1955 2000s

Ounces

McDonald's Fries Portion Sizes

Small (230 cal)

Medium (380 cal)

Large (500 cal)

Heart Attack Grill

Consequently, people are becoming heavier

• 66% US adults, 1/3 of preschoolers overweight or obese (NHANES 2004)

– Doubled over the last 20 years

• More people overweight than starving in the world (UN)

• Huge economic cost, ~$2 trillion/ yr. (McKinsey 2014)

The estimated cost of the obesity problem

McKinsey Global Institute, Discussion Paper November 201415

This is a major problem worldwide

Wall Street Journal, April 12, 2013

Obesity Trends* Among U.S. AdultsBRFSS, 1990

(*BMI ≥30, or ~ 30 lbs. overweight for 5’ 4” person)

No Data <10% 10%–14%

No Data <10% 10%–14% 15%–19%

No Data <10% 10%–14% 15%–19% ≥20%

No Data <10% 10%–14% 15%–19% 20%–24% ≥25%

No Data <10% 10%–14% 15%–19% 20%–24% 25%–29% ≥30%

Obesity Trends* Among U.S. AdultsBRFSS, 1990, 2000, 2010

(*BMI 30, or about 30 lbs. overweight for 5’4” person)

No Data <10% 10%–14% 15%–19% 20%–24% 25%–29% ≥30%

…But people are trying to lose weight

• 59% trying to lose weight (Gallup)

• How?

• Diet market size - $40 billion • Exercise market size - $18 billion

And there’s no shortage of advice

The tabloids are all over it

• “Exercise holds key to keeping weight off” Medicinenet

• “Exercise key to fighting flab: Diet takes second place in fitness stakes” Suite101.com

• “Obesity epidemic simply caused by eating too much, claims academic” Telegraph (UK)

• “Diet, not exercise, plays role in weight loss” MSNBC

• “Why exercise won’t make you thin” Time

• “Fat gene found by scientists” Times (UK)

• “The myth of the fat gene” Live Science

• “Fat gene can be trumped by exercise” Health News

So what’s a poor girl to think?

Diet… exercise… genes… Everyone arrives at some conclusion for themselves.

Their own “lay theory of obesity”

Weight gain comes from too much food…I should watch what I eat!

Weight gain comes from not enough exercise…I should go to the gym!

Which lay theories would most people hold?

• People believe that individuals can control their own weight

• People can make vast and relatively rapid changes to their diet and exercise patterns, but do nothing to alter their genetic code.

• Other factors related to obesity (e.g., sleep patterns, smoking, geographical location, etc.) influence people’s food choice or exercise patterns, but are not primary causes in themselves.

The two most commonly-held lay theories of obesity are: (1) overconsumption of food and (2) lack of exercise.

…And how does this matter?

So either she skips the cake altogether, or she lets herself eat it with a promise to burn it off with an extra workout later…

Nope, bad news!

Umm, well I just went to the gymI’ll hit the gym later…

Problem is, we’re biased

• We underestimate the calories we intake and also overestimate what we burn while exercising (Lichtman et al 1992).

• Calories? – 1110, 28g fat

• Hours of walking? – ~7

• We “reward” our exercise with additional consumption, often ingesting more calories than the exercise burned (Church et al. 2009).

And what “really” causes obesity?

• Sedentary lifestyle?• Genetics? • Eating too much?

– Americans eat 200 more calories a day than they did in 1980 (Chandon & Wansink 2007; NHANES 2004)

– Exercise rates are stagnant over this period and cannot explain obesity rate change (Westerterp and Speakman 2008)

– Gym memberships have nearly doubled 23M (1993) to 45M (2009)– 47% “regularly exercise” 1980 to 57% in 2000 (MN Heart study)

– The human genome cannot have morphed that much in only 20 years

• Eliminating one 591 ml soft drink cuts 250 calories– But…250 calories would take an hour of cycling to burn. – 250 calories a day = ~30 pounds in a year (Wansink 2006)

The American Medical Association is pretty definitive

• “Clearly, environmental causes for obesity are far more influential than genes. […]

• Obesity results from overnutrition and the primary therapeutic target is preventing or reversing overeating […]

• Exercise is associated with weight loss but its duration or intensity has minor effects on weight loss relative to diet.”

Livingston & Zylke, JAMA editorial 2012

So… our two predictions:

• Poor diet and insufficient exercise are the two predominant lay theories about the causes of obesity.

• Those who believe that poor diet causes obesity should have a lower body mass index (BMI) than those who implicate insufficient exercise.

Study 1 – These beliefs exist, and matter

• South Korean nationally representative sample (N = 254, 54% female, mean age = 32.61, range 15-68)

• Asked to indicate (in Korean) what they believed to be the primary cause of obesity: – Eating too much, – Not exercising enough, or – Genetic factors.

• Also reported their age, gender, and height and weight (converted to BMI)

Study 1 - Results

Poor diet Insufficient exercise Genetic factors0%

60%50.4%

Primary cause of obesity

Study 1 - Results

• Diet theorists had lower BMI than exercise theorists• Of the overweight people (BMI > 25), 30 (of 110) were exercise theorists,

and only 17 (of 128) were diet theorists

Poor diet Insufficient exercise Genetic factors21

Study 2 – Replicated with controls….in France• French sample (N=290)• Controls (Gutiérrez-Fisac et al 2002; NIH 2011)

– Education level (1-8) – Hours of sleep per night – Self-reported stress, anxiety, and depression (each 1=very low, 5=very high)– Underactive thyroid (hypothyroidism), Cushing's syndrome, or polycystic ovarian

syndrome (1 = yes, 2 = no) – Weight-affecting medications (corticosteroids, antidepressants, or seizure medicines;

1 = yes, 2 = no) – Current SES – Childhood SES – Home location (1=urban, 2=suburban, 3=smaller city or town, 4=rural) – Current pregnancy– Employed outside of the home – Smoker or not– Self-reported overall health (1=poor, 5=excellent) – Interest in nutrition (1=very low, 5=very high) – Self esteem

Study 2 – Continued

• Rated strength of lay theory by allocating 100 points among– Diet– Exercise– Genetics

• And an additional item “Obesity is caused more by…?” – 1 = eating too much, 7 = not exercising enough

• Uncorrelated with:– appearance self esteem – restrained eating orientation – body esteem – hyperopia– self-efficacy – impulsivity – lay theories of self control – gender

Study 2 - Results Step 1 Step 2 Step 3Age .103 .103 .116*Gender1 -.232*** -.232*** -.206***Education -.188*** -.188*** -.174***Medical conditions2 -.031 -.031 -.026Medications3 .021 .021 .026Hours of sleep per night .020 .020 .027Stress4 .032 .032 .055Socio-Economic Status (Current) -.096 -.096 -.091Socio-Economic Status (Childhood) .062 .062 .059Home location .043 .043 .054Current pregnancy5 .065 .065 .060Employed outside home5 -.025 -.025 -.021Smoker5 .145** .145** .148***Overall health -.199*** -.199** -.221***Interest in nutrition .011 .011 -.028Self-esteem .096* .096* .106*Genetics Lay Theory -.001 -.120Diet Lay Theory -.184** R2 .23 .23 .25DR2 .00 .02**

Study 2 – Results

Overweight

The effect is robust

• Seven other studies controlled for alternative possibilities– Different ways of measuring lay theories, including open-ended protocols– Priming lay theories– Introducing competing cognitions– Measuring eating behaviors– Data collected in Canada, France, HK, South Korea, and USA– And yes, doctors’ do hold the right “lay” theories

• Novel predictor of obesity – a new way of thinking about the problem

Findings regarding the existence and effects of lay theories of obesity

• People have naïve beliefs about what causes obesity– Largely implicating diet and exercise– These often diverge from science

• These beliefs guide behavior– Exercise-theorists have a higher BMI than diet-theorists– Robust to measurement

• Mere beliefs explain previously undocumented variance in BMI• Effects comparable in magnitude to many known correlates of BMI• Persist as well as explain additional variance even after these other factors

are controlled for

(Big) Food For Thought: Where Do These Beliefs Come From?

FAREED ZAKARIA (CNN): We have three times the rate of obesity as Europe. They say it is because of the snacks and fast food and high calorie drinks.

INDRA NOOYI (CEO, PEPSICO): I wish the solution was that simple. ... When I was a kid, I would come home from school, throw my bag, go out to play. My daughter comes home from school, throws her bag, goes to play, but sitting in front of the computer because their definition of play has changed. They don't go out to play. ... Lifestyles have changed.

CNN, April 17, 2011

Food companies actively promote the exercise theory

Four major ways:

• Public statements

• Lobbying

• Exercise philanthropy

• Sports sponsorships

Karnani, McFerran, and Mukhopadhyay, CMR 2014

Public statements indicting exercise and “lifestyle”

• “Balanced lifestyle” in most interviews and communications– ‘active balanced lifestyle’: Coca Cola

‘balanced active lifestyle’: McDonald’s ‘balanced and healthy lifestyle’: General Mills ‘balanced diet and lifestyle’: Unilever ‘well-balanced lifestyle’: Mars ‘a balanced lifestyle’: Nestle, PepsiCo

• Directly implicating exercise and “personal choice”– “If all consumers exercised, did what they had to do, the problem of obesity wouldn't

exist.” – Indra Nooyi, Fortune, April 27, 2010– “This is an important complicated societal issue that we all have to work together to

provide a solution. That's why we are working with government, business and civil society to have active lifestyle programs in every country we operate by 2015.” – Muhtar Kent, CEO, Coca-Cola, 2012

– “We cannot escape the role of personal responsibility we each have…. [the growth in the number of overweight people is] “mirrored by a decline in physical activity”– John Sutherland, CEO, Cadbury-Schweppes, 2004

Lobbying

• US Food industry spent $175 million on lobbying during 2009-12

• Industry-funded groups that distort research findings• e.g., The Center for Consumer Freedom (CCF) – a nonprofit “lobbying front for the

restaurant, food, beverage, and alcohol industries”: “A hefty number of studies has shown that the trend of rising obesity rates can be attributed not to increased intake of food in general (or any particular food) or to the influence of restaurants, but rather to less physical activity compounded by a variety of other factors that are constantly being explored.”

• “Of 76 industry-funded studies (March-October 2015), 70 reported results that were favorable to the sponsor’s interest” (Marion Nestle, JAMA 2016)

• Influence on guidelines, standards, and initiatives – Michelle Obama’s “Let’s Move” campaign (NYT op-ed, May 28, 2014)– US “Inter-Agency Working Group” (FDA, FTC, CDC, & Dept of Agriculture)

recommendation for voluntary standards for children’s food marketing.

Exercise philanthropy

• In-school programs, joint programs with local governments• Helping to fund and build parks, playgrounds, and gyms

• Examples:• Kelloggs’ “Get in Step” initiative encouraged people to walk an extra 2,000 steps a day

and then eat a bowl of cereal• PepsiCo’s “Smart Choices” built playgrounds in cities across the US• Kraft Foods Foundation invested more than US$ 7 million across Europe, targeting

elementary schools, sports facilities, parks and community centers. • Coca-Cola funded sports and physical activity programs in China, Brazil, Singapore, and

the Netherlands, among many others• Unilever sponsored the Flora London Marathon, the Flora Women’s 5km challenge, and

the Danish Handball Summer School.

• “If a 50-pound kid plays soccer for 45 minutes, she burns ~120 calories. But a typical post-game snack chosen from Pepsi's 'good for you' brand list – a 15.2-ounce bottle of Dole Strawberry-Kiwi juice and a reduced calorie Quaker granola bar – adds up to 320 calories.” -- WSJ

Sports sponsorships

• Coke and McDonald’s: Olympics, FIFA World Cup, UEFA soccer, multiple NBA teams

• Pepsico: NFL, Major League Baseball, Major League Soccer, ICC Cricket World Cup, Indian Premier League

• And hundreds of others

More recent data: Back to the Center for Consumer Freedom

• “A hefty number of studies has shown that the trend of rising obesity rates can be attributed not to increased intake of food in general (or any particular food) or to the influence of restaurants, but rather to less physical activity compounded by a variety of other factors that are constantly being explored.”

• Also insist: “A calorie is a calorie” – Paper titled: “An Epidemic of Obesity Myths”

Study 3: New data replicates

Overweight

And extends: Are all calories indeed the same?

Overweight

Strongly disagree

Strongly agree

In fact, if BOTH your beliefs are wrong, you’re really screwed

Believe

in diet

Do NOT beli

eve in

diet la

y theo

ry23.024.025.026.027.028.029.030.0

All calories are the sameAll calories are NOT the same

We call it “Leanwashing”

• Food marketing tries to influence people’s lay theories of obesity

• Tries to make them exercise theorists

Dual effect of “Big Food” on obesity

Policy implications: What should be done?

• “Regulate food like tobacco”? (BBC News May 2014)

• The obesity crisis is a case of market failure– Evidence of irrational behavior (Thaler and Sunstein 2015)

– Breakdown of the revealed-preference model: “Free markets… will not only provide us with what we want, as long as we can pay for it; they will also tempt us into buying things that are bad for us, whatever the costs.” (Akerlof and Shiller 2015)

• Causes of market failure– Monopoly power or lack of competition (Vickers 1995)

– Asymmetric information (Akerlof 1970)• Children; Imperfect information (leanwashing)

– Externalities (Scitovsky 1954; Greenwald and Stiglitz 1986)• Social costs of obesity; social contagion

• Therefore, need for corrective mechanisms (Datta-Chaudhuri 1990; Lerner 1972)Karnani, McFerran, and Mukhopadhyay, JACR 2016

Monopoly Power

Asymmetric Information

• Marketing to children• Problems with information disclosure• Leanwashing

Externalities •Public health expenditures•Social contagion

MARKET FAILURE

CORRECTIVE MECHANISMS

Corporate Social Responsibility

• Not incentive compatible

• Often irrelevant

Industry Self-Regulation

• Not happening• Possible

opportunity

Consumer Activism

• Sporadic• Ineffective

for many reasons

Government Intervention• Unreliable (due to lobbying)• May be the best option• But how? Taxes? Bans?

Need for

Analyzing Obesity as Market Failure

Summary of Government Intervention Remedies to Address Obesity

Policy Tool Market Failure Addressed Upside

Downside

Externalities (Public health expenditures)

Taxes at high levels could be effective (as with tobacco).

Unlikely to get political support, especially at high levels. Potentially regressive. Difficult to implement properly. Never tried at levels most likely to significantly change behavior.

Restrictions on marketing to adults

Asymmetric Information (Complexity, motivation to process, ability to comprehend)

Could have significant impact (as with tobacco).

Strong political opposition. Untested - not tried anywhere yet.

Restrictions on marketing to children

Asymmetric Information (Children as vulnerable consumers)

Demonstrable benefits. Somewhat less political opposition.

Political opposition. Patchwork of regulations.

Restrictions on distribution

Asymmetric Information (Children as vulnerable consumers)

Evidence of effectiveness. Political opposition. Hard to implement, impossible in some cases.

Product bans Asymmetric information Might work for clearly harmful ingredients.

Very unlikely to get political support unless clear evidence of harm.

Nudging Human biases in information processing and decision making

Less coercive. Possibly less political opposition. Demonstrably effective at individual level.

Difficult to scale up. Cross-context generalizability unclear.

Labeling Rules Asymmetric Information (Complex and possibly deceptive information)

Less political opposition. Unlikely to have large impact. Potentially long lag before impact.

Education

Asymmetric Information

Non-controversial.

Very mixed results. Tried before. Costly.

So... going forward

Policy implications: Firm side• Review of recent interest in sugar taxes (with Karnani & McFerran)

• Possible mechanism design problem: How to align food company incentives with the public interest?

Consumer psychology side• Attitudes to food “constituents” (with Karnani, McFerran & Das)

• What happened to those who believe equally in diet and exercise?• Does exercise really have no effect? “Diet : Weight :: Exercise : Health”?• Exercise self-reports are not accurate – need physiological data• Scope for large-scale intervention studies.

• The evolution of children’s beliefs and the effect of parents’ beliefs (with Briers & Huh)

• The effects of beliefs on BMI: Dutch LISSpanel data (with Briers & Chan)

• Lay beliefs about sleep, effects on BMI (with Weihrauch & Dewitte)

Sugar taxes

• Pre-2016• Hungary (salt, sugar, and high caffeine tax), France (soft drink tax), Mexico (soda tax),

Denmark (fat tax), Finland (soft drink tax) • Import / excise taxes on sugary beverages in Norway and several Polynesian countries • Berkeley: first American city to tax high-calorie sugary drinks (2015)

• 2016• WHO study: “reasonable and increasing evidence that appropriately designed taxes on

sugar-sweetened beverages would result in proportional reductions in consumption, especially if aimed at raising the retail price by 20% or more”

• Soda taxes approved in San Francisco, Oakland, Albany, CA, Boulder, CO, and Cook County, IL, which includes Chicago

• Expected introduction of taxes in Portugal, Spain, Estonia, UK, and Ireland• Ongoing debate in Australia, NZ, Brazil, Colombia, Philippines, Indonesia, India, and others

• Effectiveness• Berkeley soda tax: 21% drop in consumption of sugar-sweetened beverages; consumption

increased by 4% in comparison cities (Falbe et al. AJPH 2016)• Mexico: 6% decrease in purchases of taxed beverages; effect most pronounced for

households of low socioeconomic status (Colchero et al. BMJ 2016)• Pushback: American Beverage Association; Michael Bloomberg and John Arnold

Karnani, McFerran and Mukhopadhyay (invited article)

Study 4: Attitudes to GMOs and organic foods predict healthiness

Dislike organic food Like organic food2.8

Dislike GMOsLike GMOs

• N = 13,910 (Australia, HK, India, South Africa, UK, USA)• DV: Self-reported health (5-point scale)

Karnani, McFerran, Mukhopadhyay, and Das (in process)

Attitudes to GMOs and organic foods predict healthiness across countries

Australia (N=2365) Hong Kong (N=928) India (N=3786)

South Africa (N=1832) UK (N=2679) USA (N=2320)

Karnani, McFerran, Mukhopadhyay, and Das (in process)

Study 5: Attitudes to GMOs and organic foods predict BMI• Mall intercepts in North India, N = 221, 65% male, Mean age = 28, 95% at least

college educated

Thank you, and Gam Bei!

• Provides thought leadership on business and policy challenges in emerging economies

• 40+ Faculty Associates

• Founded in 2013 with support from EY

iems.ust.hk

POLICY BRIEFS

SEMINARS AND CONFERENCES

WORKING PAPERS

iems.ust.hk

Follow us on:

“HKUSTIEMS”

Subscribe to our email:

Send email to iems@ust.hk with “SUBSCRIBE” in the subject line

Mukhopadhyay iems leanwashing 02 march2017

Business