Is it only me who is driven crazy by the American Heart Association’s campaign against “added sugars”, and the attendant campaign to label foods for their added sugar content?
Look. I am no expert, so correct me if I’m wrong, but as far as I can tell from a trip around the Internet, sugar is sugar. More precisely, fructose is fructose, glucose is glucose, and so it goes. The fructose in an apple is exactly as bad for you as the fructose in a Cola drink.
Now an apple provides all sorts of good nutrients and fiber that are missing from the Cola drink. But if you want to send that message, the way to send it is to advertise that apples provide all sorts of good nutrients and fiber that are missing from Cola drinks — not to suggest (nonsensically, as far as I can tell) that the “added sugar” in the Cola drink is somehow different from the non-added sugar in an apple. If your main concern is to watch your sugar intake, the distinction doesn’t matter. If your main concern is, say, Vitamin C, then sugar counts are irrelevant anyway. If you care a little about a lot of things, then it’s good to know sugar contents, vitamin contents, and a whole lot more. But I cannot imagine any individual, in any state of the world, who is better off counting added sugar than counting total sugar.
The semi-knowledgeable people I’ve talked to tell me that it’s considered a bad idea to report total sugar content because we don’t want to scare people off eating apples, with all their offsetting health benefits. In other words, we’re not quite lying to people, but we’re twisting the presentation of facts to fool them into focusing on the wrong things. Isn’t that exactly what got Jonathan Gruber in trouble?
I’m no expert either, but my understanding is that the fiber in fruit actually slows the absorption of sugar by the body, so that it becomes a time-released energy source, rather than the cause of a short-term blood sugar spike (in which the excess sugar is more likely to turn to fat):
http://well.blogs.nytimes.com/2013/07/31/making-the-case-for-eating-fruit/
Thus it’s not just that the nutrients in an apple “make up for” the sugar; the fiber actually acts as an antidote to the harmful effects of the sugar.
Bennet, what about orange juice versus coca cola? Should both be labeled?
I can contribute! Type-1 Diabetic here.
The only caveat is that sugar actually kind of interacts with the other elements of the food in complex ways (and this is admittedly SOMEWHAT anecdotal).
If you drink a Coke you are basically drinking pure sugar. Your blood sugar levels will spike crazily then crash as your body spews insulin.
If you eat a handful of nuts, your sugar will slowly and almost imperceptibly rise over the course of a couple of hours in a way that is pretty easy to control with insulin.
If you eat a pizza, your sugar may not start rising for a bit, but then it might keep rising nearly uncontrollably for the next few hours.
So, your basic point is spot-on. People are oblivious to the fact that sugar from a “healthy” food is messing with them big-time. The small caveat (as I’ve already stated) is that sometimes the sugar behaves differently in different food “vehicles” though my guess is that science hasn’t yet fully explored/understood how/why.
A quick google search says an apple is around 10g of sugar. A can of coke 39g. Assuming a goal of watching sugar intake (and ignoring all other caveats about absorption), it’s possible that “added sugar” also means “more sugar”.
You’re of course correct on the chemistry — fructose is fructose — but Jonathan has his finger on the real issue here: generally when food manufacturers add sugar to a product, they add a lot of it, and if you’re adding a lot of sugar to something it’s usually not because it’s a food with the sort of metabolic profile that supports that level of sugar intake. (Quite the opposite, generally.) The messaging sounds weird to anyone who understands the chemistry, but it remains a good red flag for “this is something you should probably eat less of if you care about your cardiac health in the long run.”
The more interesting comparison is cola vs. fruit juice. While it’s difficult to eat enough fruit to get the same amount of sugar as a can of cola (who eats 4-5 apples in one sitting?), many people will down two glasses of orange juice. My understanding is also that the sugar is just as bad for you, and the harm from the “natural” sugar is greater than the 500% Vitamin C you get. For about 20 years I’ve been drinking heavily diluted fruit juice (10% juice, 90% water) and I prefer to eat fresh fruit.
Everyone is very bold today. (anyone tried a )?
Yes, it is presenting information in a form that is likely to achieve the desired objective, rather than presenting it in a form that conveys maximum truth about sugar levels. But sugar levels are not the point – a healthy diet is the point.
“The fructose in an apple is exactly as bad for you as the fructose in a Cola drink.”
From Bennet’s article “sugar consumed in fruit is not linked to any adverse health effects, no matter how much you eat.” In addition to this, Popkin found that “many biological factors that we noted in this article suggest that calorically sweetened beverages are associated with overconsumption when the sweetener is in a liquid form.” That is, people consume more overall if they get more calories from fructose in liquid form. So your statement is not correct; all sugar is not the same. Fructose from whole fruit is essentially harmless or beneficial, and need not be labelled. Sugar in drinks may pose a particular threat.
As Ricardo Cruz and Jack PQ say, what about juice?
The picture here is that this could be just as harmful as fructose from added caloric sweeteners (high fructose corn syrup mostly) to soda, although there is some dispute (particularly juiced whole fruit smoothies are better than filtered juice). So there does seem to be a good case for labelling sugar from this source.
A possible objection to this is that this would put people off fruit in general – not very convincing. Another might be that fruit juice tends to be consumed in small portions that are not harmful, and reducing these might unintentionally end up with more soda consumption. That is speculation on my part.
Overall there seems to be no good reason for labelling added sugar in drinks and not natural sugar in juice.
As for added sugar in other foods – it is probably difficult to over-consume sugar if you avoid added sugar and obvious sweet things (cakes, chocolate etc). We have seen that sugar in fruit is essentially harmless, so total sugars would be misleading. Added sugar may be the best label to allow people to maintain a healthy diet.
So my prescription: label total sugars in drinks and total and added sugars in solid food.
My mother, who is diabetic, can’t believe that juice is bad for her because the sugar is “natural”. She’ll drink a glass of juice (30g sugar) but won’t put one sugar in her coffee (4g sugar).
I’d argue that the presentation of facts to emphasize “added sugar” is HURTING my mother’s health.
Every dietician that I’ve asked, has said that fruit juice is bad for your overall health.
This is counter-intuitive at first, because it seems that if an apple is healthy, then any subset of what’s in an apple must be good for you as well, or at least not bad. The flaw in this thinking is that apple juice is missing the fiber in the apple that counteracts the sugar.
Whole-fruit smoothies are fine because they include the fiber in the original fruit.
This is not even counting the fact that an actual fruit (or a whole-fruit smoothie) also helps you feel full, and less likely to eat junk.
Steve (and Harold) raise a fascinating issue about language and communication:
This is a famous problem in policy design, especially involving public health. Imagine three scenarios.
1. I present data regarding the hazards of smoking to the American public in Latin. They don’t change their smoking behavior.
2. I present data regarding the hazards of smoking to the American public in English. They don’t change their behavior.
3. I present an ad depicting cigarette executives cackling gleefully about how they’ve duped people into thinking that suicidal behaviors such as smoking are “cool.” People quit smoking.
Is Scenario 2 more like Scenario 1 or Scenario 3? There is plenty of reason to think that presenting info in Latin to people who don’t speak Latin will be ineffectual. Likewise, there’s plenty of reason to think that presenting info in dispassionate scientific language to people who have, since the dawn of language, learned to weigh the salience of a message based on the passion with which it is presented, be equally ineffectual.
Assuming I read the data to lead to the conclusion that most people would want to quit smoking, arguably Scenario 3 reflects the best translation of this data into the vernacular of the realm.
I don’t think we can entirely trust typical studies of nutrition because they are not randomized. Say you sample 1000 people who eat a lot of fruit and do not drink cola, and compare their health to 1000 people who drink a lot of cola and do not eat fruit. It would be wrong to attribute health differences to fruit vs. cola. Fruit-eaters are probably different in many other ways (wealthier, more educated, exercise more).
#10 – I don’t know, in some ways this seems harder to justify than pure command and control – either way I review the facts and decide what’s best — not just for me but for everyone (without, of course, a similar level of detail about their personal situations), rather than allowing them the privilege of making an informed objective decision; but for some reason (maybe easier to maintain the control?) it’s better to trick them into thinking they chose it for themselves (for the wrong reasons)? Pretty cynical, even when an economist does it.
As Walton Hale Hamilton remarked, “Business succeeds rather better than the state in imposing restraints upon individuals, because its imperatives are disguised as choices.”
To be sure, ideally we pursue Scenario 2 in any event; the real question is whether we (and in particular, government) also pursues Scenario 3. Does an ad campaign reflect a means of translating data to make it meaningful to people who are not eggheads – or is it merely propaganda that is pretty much unrelated to the data?
Imagine that conservative lobbying groups find that Hillary’s tax plans would impede economic growth. And imagine this prompts them to fund ad campaigns attacking Hillary — for failing to save American lives in Bengasi. It would require quite a stretch of credulity to say the ad campaign was a meaningful exposition of tax policy. But is it really so different than the suggestion that an anti-smoking campaign featuring cackling tobacco execs is a meaningful exposition of health data?
(This is not an entirely hypothetical question for me. My job involves briefing superiors. I have some idea about which issues my superiors will care about and which they won’t. If I brief them, and they proceed to act as if they never heard my briefing, have I failed in my job? In other words, if I fail to discuss issues in a manner that will be meaningful to my superiors, have I basically given a briefing in Latin? On the other hand, if I measure my success on the basis of how much my superiors conform their behavior to my views, am I creating an incentive to manipulate them and usurp their authority?)
“As Walton Hale Hamilton remarked, “Business succeeds rather better than the state in imposing restraints upon individuals, because its imperatives are disguised as choices.”
Indeed. I don’t see why Govt. should be prevented from using techniques ubiquitous in the private sector.
“rather than allowing them the privilege of making an informed objective decision” This is a myth. Nobody really makes an informed objective decision ever because we cannot be totally objective. That is why nearly all advertising is not just a list of facts.
Smoking is an interesting case. In the 1950’s UK there was one of the first large epidemiological trials on smoking – called the British Doctors Study. The results clearly showed that smoking was bad for you. Doctors gave up smoking much more than everyone else. However, it is believed that this is not because they were better able to analyse and interpret the results, but because they could relate to the subjects more as fellow doctors. One of the authors (Peto) said “”A lot of the people in this study stopped because they read the BMJ and said, ‘Bloody hell, this doesn’t just kill patients, it kills doctors too.’ ”
There are two separate issues. If we tell people about only added sugar, do we get a result more in line with what the AHA wants? An example, apples are good for you, coke is not, so advertising total sugar would be true, but would end up with the “wrong” result (reduced apple consumption). It seems likely that with juice / coke this is not the case, and advertising added rather than total sugar will not end up with the “right” result.
The other issue is whether bodies like the AHA should present the data in a form that results in people acting like the AHA wants, even if it is not the whole truth. So even if advertising total sugar in apples will inevitably result in reduced apple consumption, should the AHA do so anyway?
Maybe I’m being overly moralistic or paternalistic, but I see overall expected sugar content as a public good.
If people’s palates and metabolisms expect to eat more sugar, then their diets will be far less healthy.
An individual food producer (say, of frozen pizza) can make his product more successful by adding sugar. He realizes a private gain, but as a result, people get even more used to high sugar, and they come to expect it across all products. Sounds like a negative externality to me.
The crime isn’t sugar in goods. Like you said, you don’t want to “punish” apple farmers or discourage apple consumption. You want to punish and discourage food producers from adding sugar. That’s the externality which should be taxed through increased information.
You’re too kind.
Of course the difference is companies don’t have the power of force behind their ‘impositions’. While they can face competing messages. Once we trick people into approving a law or regulation they’re no longer free to change their minds. Reasons to tread more lightly. I realize in this case we’re just talking about labeling, but this still leads to…
“nobody really makes an informed objective decision ever”
And yet we always seem to assume that we alone can decipher the ‘right’ action. If no one can be objective, what is left of the case for tricking others into taking a preferred course…or even presuming there are any ‘right’ answers? Even if we were homogenous? To me good economics has to allow for different *and valid* choices from the same set of facts about costs and benefits. Even for something like smoking. How you like them apples?
Sugar is sugar but it matters a lot how quickly ingested sugar gets into the bloodstream. Sugar in drinks hits the bloodstream in a shorter period than the same amount of sugar in an apple. This can cause a spike in blood sugar. There is, metabolically, an important difference between a surge and a drip, even if the same amount of stuff is involved.
Yes and very few people in the USA are deficient in those other nutrients. Sugar is good food as long you do not consume too many calories.
I’m willing to give AHA a pass on their added sugar obsession for this reason: AHA has taken a market approach to promoting heart healthy eating with their private food certification program. Far better than Nanny-state mandates enforced by bloated bureaucracies.http://www.heart.org/HEARTORG/GettingHealthy/NutritionCenter/Heart-CheckMarkCertification/How-the-Heart-Check-Food-Certification-Program-Works_UCM_300133_Article.jsp#mainContent
#15 -yes, you are the only one.
#16 “And yet we always seem to assume that we alone can decipher the ‘right’ action.” It is not based on a whim. It is based on research. There is a huge difference between saying I like apples so everyone should eat apples and saying research has shown that if you link apples to sugar there will be reduction in apple consumption, even if you explain that the sugar in apples does you no harm. No decision is ever totally objective, but some are more so than others.
To be sure. So we assume we alone will take the time to digest research findings (and then act objectively).
By all means let’s disseminate facts. Took me 30 seconds on this blog to learn about potential metabolic differences between fruit and juice – a more credible message that’s more likely to alter my behavior than some ‘red sugar / blue sugar’ argument.
People can pick up on spin, and being Gruber-ed erodes trust in the messenger. It also doesn’t help that recent history includes misguided ‘facts’ about fat, eggs, infant formula etc.
Why dont you guys defend labels warning against sugar from juice rather than muddling the issue with scary labels which only make people skeptic of such campaigns?
“So we assume we alone will take the time to digest research findings (and then act objectively).” No, if we are wishing to send out an effective message we do not assume anything. We look at or conduct research and discover what messages are likely to be effective. We do not assume anything about ourselves in relation to the behavior in question.
R11/ All my research involves laboratory experiments (of human economic behavior). This allows me to completely randomize across treatments. So, I have no experience in attempting to draw conclusions from naturally-occurring data. However, I do read business/economics research that deals with the sorts of issues that cause you concern about nutrition research: essentially the subjects have self-selected their treatment. I must imagine anyone doing nutrition/health/disease research (certainly that published in respected, refereed journals) is well aware of the issue you address and has attempted to control for such factors; they certainly must regularly attempt to address that very point. Still, I’ve seen business/economics research that does a poor job on that dimension. Indeed, I suggest the real work in analyzing naturally-occurring data is controlling for self-selection.
I don’t know about you but I’m not doing any original research here. So by having this debate we’re assuming *as consumers* you and I can understand a factual message and the appropriate behavioral response. So why do we assume others must be spun? (Gruber actually has a better case for playing on the rational ignorance of voters – here we’re talking about what people choose to put into their bodies.)
BTW saying we’re not assuming anything about our own behavior suggests even this understanding doesn’t necessarily lead us to the proper response? That doesn’t inspire much confidence in the spin sans understanding – unless the lie is pretty brazen (e.g. “fruit juice causes cancer”). Which probably destroys the institutional credibility pretty quickly.
Yes, it matters how (or really in what) the sugar is delivered for sure. Really it’s best to have some protein, carbs, fat, and fiber in every meal. The main theory is that if you constantly eat refined foods that make your blood sugar spike quickly , you can wear out your insulin receptors faster than someone who eats a more balanced diet.
“So why do we assume others must be spun?” I dunno, perhaps because we have eyes ears and a brain? (By the way, it is not just others, I include myself in this. I know my preferences are affected by the message, not just the facts.)
How many advertisers take your line “assuming *as consumers* you and I can understand a factual message and the appropriate behavioral response.”? None, which is why adverts are not simply lists of facts.
The AHA would be foolish to ignore all this evidence and assume that people will respond as they would wish.
“…respond as they would wish.” And that’s the issue — what is it exactly that would justify their tricking people into behaving a certain way?
With business advertising there’s no pretense of pursuing some (more) objectively “right” answer in the name of the public good. And of course a business doesn’t have the legal power of force. So my question was: if we really think “we” have “the answer” — for everyone? – doesn’t it seem cleaner, maybe even more noble, to just ban certain types of food processing rather than deceive people with clever labels? (As I said deception may be the path of least resistance in the SR but risks longer-term credibility). Of course the less objective or universal “the answer” is, the stronger the case for just giving us the facts (and maybe having us bear more of the risks of our chosen behavior). Some might think this is more consistent with liberty as well.
” to just ban certain types of food processing rather than deceive people with clever labels?” We do ban lots of types of food processing – those that would leave the food toxic or contaminated with bacteria for example. We must strike a balance.
There are interesting aspects to this sort of paternalism. Say you run a cafeteria. If you do some research and find that people take more apples than cakes if you put the apples first, or more cakes than apples if you put the cakes first. Each costs the same and gives you the same profit. Should you put the apples or the cakes first? Should it concern you at all? What if your children ate at the cafeteria?
Personally, since I must put one of them first, I would put the apples first.
I agree we can presume people wouldn’t knowingly choose to ingest things that are imminently life-threatening. To me the sugar example is more “we think this is bad for you” = a value judgment; again an informed person could decide the current pleasure is worth some abstract, probabilistic future risk, so IMO we should make the case fairly and accurately. To me the owner of a cafeteria is generally free to impose whatever values she wants or be neutral with respect to her customers’ preferences (it’s not her fault if “we” choose to subsidize people’s health choices). In any event she might want to rotate them so apples don’t go bad and cakes don’t get stale. I’m charged with being involved in what my kids are eating.
“so IMO we should make the case fairly and accurately” Labelling added sugar is accurate.
did u read the post? :)
#33 I did read the post. We are not arguing about the accuracy of the statement “This has Xg of added sugar” It is an accurate statement. You saying that you believe labelling should be accurate does not help decide if we should label added or total sugar.
You’re right – we’re not arguing about that. “*Fairly and* accurately” technically accurate but misleading, e.g. which can make people think soda bad, fruit juice good. I would have said something like “as objectively as possible” but figured you would object to that.
Indeed I may object, as facts are facts, and what is misleading is precisely the point. It may mislead people into thinking that apples are bad for them if we label the sugar content. Labelling the added sugar will not be a problem in this case. So added sugar is both fair and accurate for apples. For fruit juice the opposite is probably the case. I suggested labelling only total sugar for drinks and added sugar for solid food as a compromise. If we had to stick to only one label, which would be fairer -misleading people that apples are bad or misleading people that fruit juice is not bad? As we greed, both added and total sugar statements are accurate.
I’m glad we agree it’s best not to mislead people
Yes, always best to finish with agreement.
“I am no expert”
When has that ever stopped you from acting as if you were?