Should DC Tax Soda?

Earlier this month, the DC City Council passed the Healthy Schools Act, which will raise nutritional standards for school meals, increase the amount of physical and health education students receive, create school gardens, and do all kinds of other commendable things. The difficult part is that it'll cost $6.5 million annually, and we're in the middle of a budget crunch.

Councilmember Mary Cheh has proposed a funding mechanism that has the potential to not only raise money but to fight obesity at the same time: a one-cent-per-ounce excise tax on sugar-sweetened beverages. (The beverages are non-alcoholic ones containing "caloric sweeteners," and the definition excludes milk products and beverages that contain more than 70% juice by volume; draft legislation is here.) The first $6.5 million raised each year will fund implementation of the Healthy Schools Act, and funds beyond that will go to the Healthy Food Access Initiative, the Healthy Corner Store Program, and other worthy-sounding projects. The rationale for this kind of tax is pretty solid.

Taxing a product is often a good way to get people to buy less of it -- for instance, jacking up tobacco taxes has helped reduce smoking. The money raised by such a tax can also be used to offset the toll of the related behavior. The millions of vehicles driving in the US take a toll on our roadways, so we tax gasoline to fill the Highway Trust Fund that pays for roadwork (or much of it, anyway). Sweetened beverages contribute to obesity, and this tax would fund efforts aimed at decreasing obesity.

Effects on Obesity
How much would the tax affect obesity rates? In a 2009 New England Journal of Medicine article, Brownell et al estimate "an excise tax of 1 cent per ounce would lead to a minimum reduction of 10% in calorie consumption from sweetened beverages ... a reduction that is sufficient for weight loss and reduction in risk." They base this estimate on research into price elasticity (how much demand will change with a change in price), and assume that in some cases people will substitute other forms of calories for sweetened beverages they don't drink.

In a recent study published in Health Affairs, Sturm et al combine data from the Early Childhood Longitudinal Study - Kindergarten Cohort with information on soda tax rates in different states. Several states already tax soda at higher rates than they do other food or beverage items, with the average differential being 3.5%. The results of their analysis suggest that these relatively small taxes are unlikely to measurably affect children's soda consumption overall -- but they may affect consumption by subgroups who are at particular risk for obesity, including children in low-income families.

A one-cent-per-ounce tax will raise the price of sweetened beverages more than existing state soda taxes already do -- between 10-20%, depending on the original price. (If a 12-ounce can of a sweetened beverage originally cost $1, its price will rise to $1.12, a 12% increase. If a one-liter bottle originally cost $1.69, its price will rise to $2.03, a 20% increase.) Sturm et al estimate that an 18% differential soda tax would correspond with a 20% reduction in excess BMI gain. "No other anti-obesity policy has demonstrated a reduction of that magnitude yet," they note.

So, Should We Do It?
The price increase under consideration is pretty large in percentage terms, but it's still just 12 cents for a can of soda or 34 cents for a one-liter bottle. Many soda buyers won't notice the difference, but low-income soda consumers will. Obesity rates are highest in DC's poorest neighborhoods -- in Wards 7 and 8, which have the lowest median incomes, 40% and 42% of residents (respectively) are obese -- so it's arguable that the poorest residents most need to feel the effect of the tax and change their purchasing habits accordingly.

I'm not entirely comfortable with the idea that we can implement a policy that will shape poor people's behavior just because they're poor. (Of course, many of our policies end up having that effect, but that's usually not their intent.) I'd much rather see us change the things that allow sweetened beverages to be so cheap in the first place, like the billions of dollars in subsidies for corn (and, by extension, corn syrup) and the tax breaks for advertising. However, I suspect that beverage companies could lose these subsidies and still keep their prices close to what they are today -- and they'd might well decide to do just that in order to preserve their market share.

What I'd really like to see is for our country to address obesity by promoting health in a comprehensive way: Make healthier foods more affordable; improve health education; shift from car dependence to more walking, biking, and public transportation; and make it easier and safer for people to exercise in their communities. Policies to move us in this direction will require massive amounts of time, money, and other resources, though.

In the end, I have to support DC's soda tax proposal because it's likely to be effective and because we can implement it quickly. What do you think?

More like this

On the question of whether a soda tax can actually reduce the amount of sugary drinks people consume, a new study finds the resounding answer is “yes.” In November 2014, Berkeley, California, voters passed the nation’s first tax on sugar-sweetened beverages in an effort to reduce their impact as a…
Another day, another study that shows soda taxes work to reduce the consumption of beverages associated with costly chronic diseases in children and adults. This time it’s a study on Mexico’s sugar-sweetened beverage tax, which went into effect at the start of 2014 and tacked on 1 peso per liter of…
In ongoing public health efforts to curb the obesity epidemic, better menu and nutrition labeling is often tapped as a low-cost way to help make the healthy choice, the easy choice. And while the evidence on the effectiveness of such interventions is still emerging, a recent study found that…
Five million dollars. That’s how much the fast food industry spends every day to peddle largely unhealthy foods to children. And because studies have found that exposure to food marketing does indeed make kids want to eat more, advertising is often tapped as an obvious way to address child obesity…

I've never supported taxing soft drinks for two reasons: the import tariffs on cane sugar and farm subsidies for corn. I find it quite absurd and inefficient to take tax monies and subsidize corn then tax a product made with that corn in the form of HFCS. The price of soft drinks does not reflect the true costs in making it so that should be corrected first before taxing.

Assuming that taxing sugary drinks would accomplish the intended purposes (simultaneously raise money for schools while improving publich health), it is still a bad idea for the District.

As a tiny city-state sandwiched between two other jurisdictions, D.C. just can't get away with excise taxes on anything.

Several years ago D.C. raised taxes on gasoline. In response, local residents drove 15 minutes away to Virginia or Maryland to purchase fuel at cheaper rates.

Which means that an excise tax on soda would be easily evaded by any person who can easily cross the border.

Meaning that the tax would be superbly regressive.
Which I have it on good authority is considered a bad thing.

By Anonymous (not verified) on 17 May 2010 #permalink

It excludes milk products?

So chocolate milk, which already has more sugar than soda, will not be taxed and thus parents will be given even more incentive to buy it. Nice, push people to choose a worse option, that's smart. That's exactly what will happen: kids will cry for sweet drinks, when parents give in, instead of buying soda they'll buy cheaper chocolate milk thinking "it's milk it's better for them" when the sugar content is so high it's actually worse.

Did you see Jamie Kennedy try to get the chocolate milk out of the schools? It's impossible to fight the government who "knows best" even when they are pushing us to make horrible choices.

By plutosdad (not verified) on 17 May 2010 #permalink

Anonymous: Whether people would just shift their soda purchases outside DC is definitely a consideration. According to Mary Cheh, Maryland taxes sugar-sweetened drinks at 6%, and Virginia has both an excise tax and a 1.5% sales tax on them. DC's tax would still be higher, but we might be talking an extra dollar or two per shopping trip. Is it worth the extra time and money (for gas or fares) to save that much? Going outside DC to buy groceries doesn't seem like a rational choice for anyone who doesn't already live very close to the border - but of course we can't assume everyone behaves rationally.

plutosdad: I'm guessing that chocolate milk is excluded because it contains calcium and other nutrients. But you're right that its sugar content is high and it's not a good substitute for soda.

Neighboring state (from where I am) NY is considering similar legislation.
Silly in my opinion. Will diet drinks bear the tax? What about excluding apple juice? That contains minimal nutrition (maybe some vitamin C) but loads of sugar (FTR, never gave it to my kids; water instead).
Better to educate people to teach their children to make smart food choices.

DC's legislation won't cover diet sodas, so the hope is that some people will switch from regular to diet drinks and consume fewer calories.

Apple juice will be exempt from the tax as long as the drink is 70% juice. Like chocolate milk, it may not be the best choice, but its nutrients make it an improvement over soda.

By Liz Borkowski (not verified) on 18 May 2010 #permalink

Absolutely not!

34 pence a litre doesn't sound like much. But that's only the beginning.

Taxes increase, and they broaden in scope.

When you can't buy an oil-free, fat-free, salt-free, vinegar-free, taste-free organic rocket salad without paying some sort of tax on it, maybe then you'll see why it's a bad idea.

Hi Liz, it's Julia! The diet soda exemption gives me pause. Is it really healthier to consume artificial sweeteners instead of HFCS? Obviously, one would consume fewer calories that way, but what about the other health concerns associated with artificial sweeteners? Excess calorie consumption is only part of our general nutrition problem. Calories and fat people are an easy thing to focus on, but if the point of the law is to encourage better beverage buying behavior, I'm not convinced this will do the trick. Seems like the diet soda aspect of the proposed legislation could just be trading one bad thing for another.

By bindicated (not verified) on 18 May 2010 #permalink

Hi, Julia! I think there are a few issues here:

1) The goal is to replace the worst beverages with ones that are less awful - chocolate milk, apple juice, diet soda - even if they're not the best choices.

2) Most people who read mainstream newspapers and magazines have probably heard over and over again that obesity is a major health concern and that sugary drinks contribute to it. By contrast, I doubt that anyone who doesn't deliberately seek out information on the latest health research can explain what health concerns exist about artificial sweeteners and how strong the evidence is. (And I *do* seek out health research info, but I still can't tell you offhand what the research says about artificial sweeteners.)

3) Even if Councilmember Cheh thinks there are problems with diet drinks, including them in the tax proposal would make it an even harder sell. Right now, the many people who drink only diet sodas won't be much affected by the tax, and beverage manufacturers can replace some lost sugary-drink sales with sales of diet alternatives. If that changed, opposition would increase.

By Liz Borkowski (not verified) on 18 May 2010 #permalink

Ideally, I'd like to see other policy reforms, but given what people I know who teach in DCPS say about the need their kids have for more PE time, I think we should do it.

Regressive taxes on things you have to buy are bad. Regressive taxes on things that are pure choice bother me less.

It's also import to note that DC voters aren't in a position to change federal policy on anything so we need to use the tools we have.

By katydid13 (not verified) on 19 May 2010 #permalink

The discussion is just about the pros and cons of the tax itself and its effects on soda consumption, as well as more peripheral issues such as chocolate milk. However, it should be noted that the effects of soda consumption and/or chocolate milk consumption on childhood obesity are only rather weakly documented to begin with. And perhaps more importantly, there are virtually no intervention data showing that if children drank fewer sodas or less chocolate milk, their weights would change. They might just eat something else instead, you know. The whole field of childhood obesity interventions seems to be full of "good ideas" or "any reasonable person knows ..." ideas that either have not been put to the test or when tested have failed to affect the outcome of interest. Listen to the Freakonomics podcast on obesity where they ask Zeke Emmanuel if these interventions will work. He doesn't want to say "No" because then what would be the point. But he also doesn't want to say "Yes" because in fact we don't really know. So he uncharacteristically waffles (no pun intended).

It behooves public health people to be critical about these things and not just take them at face value because they seem like such good ideas or because you think they are opposed by corporate interests or whatever.

Carolyn, I definitely wouldn't support the tax if I didn't feel that there were compelling evidence linking lower soda consumption with lower weight. Brownell et al give a nice summary of some of the research (I'm pasting selectively - the whole section is worth a read):

In recent decades, intake of sugar-sweetened beverages has increased around the globe; for example, intake in Mexico doubled between 1999 and 2006 across all age groups.8 Between 1977 and 2002, the per capita intake of caloric beverages doubled in the United States across all age groups9 (Figure 1). The most recent data (2005â2006) show that children and adults in the United States consume about 172 and 175 kcal daily, respectively, per capita from sugar-sweetened beverages.

The relationship between the consumption of sugar-sweetened beverages and body weight has been examined in many cross-sectional and longitudinal studies and has been summarized in systematic reviews.1,2 A meta-analysis showed positive associations between the intake of sugar-sweetened beverages and body weight â associations that were stronger in longitudinal studies than in cross-sectional studies and in studies that were not funded by the beverage industry than in those that were.2 A meta-analysis of studies involving children10 â a meta-analysis that was supported by the beverage industry â was interpreted as showing that there was no evidence of an association between consumption of sugar-sweetened beverages and body weight, but it erroneously gave large weight to several small negative studies; when a more realistic weighting was used, the meta-analysis summary supported a positive association.11 A prospective study involving middle-school students over the course of 2 academic years showed that the risk of becoming obese increased by 60% for every additional serving of sugar-sweetened beverages per day.12 In an 8-year prospective study involving women, those who increased their consumption of sugar-sweetened beverages at year 4 and maintained this increase gained 8 kg, whereas those who decreased their intake of sugar-sweetened beverages at year 4 and maintained this decrease gained only 2.8 kg.

... Four long-term, randomized, controlled trials examining the relationship between the consumption of sugar-sweetened beverages and body weight have been reported; the results showed the strongest effects among overweight persons. A school-based intervention to reduce the consumption of carbonated beverages was assessed among 644 students, 7 to 11 years of age, in the United Kingdom with the use of a cluster design.16 After 1 year, the intervention group, as compared with the control group, had a nonsignificantly lower mean body-mass index (the weight in kilograms divided by the square of the height in meters) and a significant 7.7% lower incidence of obesity. In a study involving 1140 Brazilian schoolchildren, 9 to 12 years of age, that was designed to discourage the consumption of sugar-sweetened beverages, no overall effect on body-mass index was observed during the 9-month academic year.17 Among students who were overweight at baseline, the body-mass index was nonsignificantly decreased in the intervention group as compared with the control group; the difference was significant among overweight girls. In another clinical trial, 103 high-school students in Boston were assigned to a control group or to an intervention group that received home delivery of noncaloric beverages for 25 weeks. The body-mass index was nonsignificantly reduced in the overall intervention group, but among students in the upper third of body-mass index at baseline, there was a significant decrease in the body-mass index in the intervention group, as compared with the control group (a decrease of 0.63 vs. an increase of 0.12).18 The effects of replacing sugar-sweetened beverages with milk products were examined among 98 overweight Chilean children.19 After 16 weeks, there was a nonsignificantly lower increase in the percentage of body fat in the intervention group than in the control group (0.36% and 0.78% increase, respectively), whereas there was a significantly greater increase in lean mass in the intervention group (0.92 vs. 0.62 kg).

This is not to say that reducing soda consumption will either work for everyone or solve the obesity problem on its own. As you say, some people will replace sweetened-beverage calories with calories from other sources - in their calculation, Brownell et al assumed that substitutions will occur for 25% of the calories.

By Liz Borkowski (not verified) on 20 May 2010 #permalink

I have read the Brownell article. It's a lot of probablys and maybes and ecological analyses. It does mention four specific intervention trials. None had a significant impact on body weight, although various subgroup analyses could be interpreted as possibly showing something. Those subgroup analyses are really hypothesis generating, but none of those hypotheses have been tested and to the extent that Study A actually is inconsistent with the hypothesis from Study B, they don't really support each other. For instance the Brazilian study showed no significant effect overall and suggests the effect is among overweight children,but the Chilean study was only among overweight children and also showed no significant effect.

It is hard to see that major public health policy should be based on something for which the only extant intervention trials show no significant effect. I don't think people would be so accepting of this if the topic were something less culturally sensitive. Let's face it,a lot of middle class health professionals think cappuccinos are fine but soft drinks are culturally objectionable. Suppose we had four intervention trials of a drug to prevent diabetes and none of the effects were significant - would people infer that therefore everyone should take the drug in order to prevent diabetes?

I don't much like soft drinks myself and believe me I am not in the pay of anyone who would be financially affected by this one way or the other. I would be perfectly happy if children drank less soda. I just am struck by how selective the interpretation of the evidence here seems to be.

Washington State (the other Washington) just passed a temporary tax (2 cents per 12 ounces) on carbonated soft drinks, starting July 1. It's one of several new taxes aimed at generating revenue to offset proejcted state budget deficits. It'll be interesting to see what effect it has on soda consumption, if any.

Carolyn, you make a good point about the subgroups - part of the reason I find the results of the studies encouraging is that they do find the greatest effects in the most vulnerable subgroups. It would be good to see more research starting from hypotheses specifically about subgroups.

The challenge in public policy is always to know when there's enough research, and to not delay a potentially useful intervention for years and years in an attempt to amass an unassailable mountain of data. If this were a diabetes drug trial, I probably would want to see more and larger studies, because drugs affect people's bodies in complex ways that aren't always evident in smaller, shorter trials.

In the case of sugar-sweetened beverages, we know they're unhealthy, we know that taxes on tobacco (and, I believe alcohol) have successfully reduced consumption, and we have research data that suggests a soda tax can reduce obesity. Taken together, I think it makes a strong case for a soda tax.

By Liz Borkowski (not verified) on 24 May 2010 #permalink

Perhaps you know more about these studies than the Brownell article mentions. But as I read it, the UK study did not show an effect. The Brazil study also did not show an effect. And the Brazil study also did not show an effect in the overweight children. But then it did in the sub-subgroup of overweight girls. The Boston study did not show an effect, but it did in the subgroup of the heaviest third. And the Chilean study was limited to overweight children and did not show an effect. So overall, none of the studies showed an effect. And the one that was limited to overweight children didn't show an effect.

This is not a difficult intervention. Don't you even wonder why there is no intervention evidence supporting this? Don't you think that should at least be something to consider before just jumping on the bandwagon? The question is not how much evidence do you require, but how little evidence would be too little? They've tested this. So far it doesn't seem to work. Don't you even think it's possible that that's because it actually doesn't work? Shouldn't that be at least something to consider?

I'm not surprised to see that the only significant results were for subgroups. It's hard to measure whether reducing soda consumption will reduce obesity, because it's hard to get people to reduce their soda consumption; measure how much soda they're really consuming; and follow them for long enough to see potential changes in body mass, which usually occur over several months or years. I liked the Boston study because it actually gave participants beverages to drink as a replacement; the duration of 16 weeks was very short for seeing an impact on BMI, though. Since it was a pilot study, I hope we'll eventually be seeing the design used on a larger group for a longer time.

At the moment, though, we DC residents have been given a choice between the alarming status quo and a policy that would potentially reduce obesity and definitely fund healthier foods in schools. It's certainly is possible that the tax will only provide funds for a worthwhile program and not reduce overall obesity in any measurable way (although I would consider it a success even if it only affected those who are already obese). I think there's enough evidence (not just on interventions, but on the effects of taxes and the contribution of soda to obesity) to warrant adopting the policy.

By Liz Borkowski (not verified) on 25 May 2010 #permalink

Check out the new JAMA article on "spin" in randomized trials and you will see that the way these interventions are being interpreted is an excellent example of the spin that they are talking about, where a trial fails its primary objective but is nonetheless interpreted as positive. Public health should try to avoid this kind of spin. ("Reporting and Interpretation of Randomized Controlled Trials With Statistically Nonsignificant Results for Primary Outcomes").

The trials are not about "reducing obesity." They are about changing BMI. It didn't change. So on the basis of the evidence available, decreasing soda consumption won't have any effect on BMI. Yes, you may argue that somehow it will decrease BMI in obese kids but the only study conducted specificallly on overweight kids apparently does not support that view.