Friday, October 10, 2014

THE BERKELEY SODA TAX: ECONOMICS ON THE BALLOT


In this world, nothing can be certain except death and taxes. – Benjamin Franklin


Berkeley made the front page of the Oct. 7 National section of The New York Times. Why? Because the November Berkeley city ballot will contain Proposition D - A City of Berkeley Sugary Beverages and Soda Taxthat, if passed by voters, will add a 1-cent per fluid ounce tax on distributors of certain drinks sold in the city. The additional tax on one, 20-oz bottle of Coke or Mountain Dew will be 20₵, which represents about 11% of the current pre-tax retail price at a grocery store.  

The stated purpose of this proposition is: "… is to diminish the human and economic costs of diseases associated with the consumption of sugary drinks by discouraging their distribution and consumption in Berkeley through a tax. Specifically, the purpose of this ordinance is to tax the distribution of sugary drinks and the products used to make them."

A growing list of localities has attempted, so far unsuccessfully, to implement specific taxes on sugar-sweetened beverages (SSBs), which include soda. This list now includes San Francisco and Berkeley that have each placed different tax propositions for SSBs on their November ballots. The debate about the tax is already bubbling up in Berkeley.

I discuss here economic information and analysis results that can be relevant in assessing what the economic effects of Prop D will be. To be clear, people decide to vote for or against a tax proposition like this one on the basis of many factors. Such factors can include thoughts about socio-economic equity and improving public health, to name but two that transcend economics. But here's the economics.

Like many tax-oriented propositions, Prop D is intricate, citing what beverages and distributors are subject to and exempt from the tax and what entities will directly pay for the tax. The language of Prop D includes the following:

"Shall an ordinance imposing a 1¢ per ounce general tax on the distribution of high-calorie, sugary drinks (e.g., sodas, energy drinks, presweetened teas) and sweeteners used to sweeten such drinks, but exempting: (1) sweeteners (e.g., sugar, honey, syrups) typically used by consumers and distributed to grocery stores; (2) drinks and sweeteners distributed to very small retailers; (3) diet drinks, milk products, 100% juice, baby formula, alcohol, or drinks taken for medical reasons, be adopted?"

The website of a major proponent of Berkeley's Prop D, Berkeley vs. Big Soda, states "The link between sugary drinks and diseases like diabetes is undeniable, and 40% of kids will get diabetes in their lifetimes unless we do something about it." The proponents argue that the tax will reduce the incidence of obesity and diabetes, especially in kids.

Unlike the soda tax on San Francisco's ballot, the Berkeley soda tax revenues will be put into the city's general fund, and perhaps spent on new or expanded health programs in the city at the sole discretion of the City Council, as advised by the proposition's newly-formed Panel of Experts. The fact that the resultant soda tax revenues will not necessarily fund such health programs is something the opponents prominently mention.[1]

I will address 3 questions about Prop D's proposed tax. (1) Who will pay the Berkeley soda tax? (2) How much will the tax reduce the consumption of SSBs? And (3) Can soda taxes reduce obesity and diabetes?

1) Who will pay the soda tax?  The "incidence"[2] of an excise tax like one based on sales of certain SSBs is determined, in economic terms, by the price elasticity of demand (PED) and price elasticity of supply (PES) of the taxed good. The more inelastic the good's PED (the more insensitive consumers' demand is to changes in a good's price) is and the more elastic is its PES (the more sensitive producers are to making more of the good as price increases), the greater proportion of the tax is paid by the ultimate consumer rather than the producer (or in Prop D's terms, the "distributor"). Using the information cited below, Berkeley consumers of soda are very likely to be paying more for SSBs should the tax be approved, despite what city officials have said.

I could find no specific information about the value of SSB producers' PES. Nevertheless, it is logical to assume that the PES for SSBs could be moderately elastic (PES> 1.0) since key inputs are readily available (e.g., water, "secret" soda concentrate-syrup) and marginal input and production costs are fairly low.

The second important measurement for assessing Berkeley's soda tax incidence is the value of SSB consumers' price elasticity – their price elasticity of demand. I found references to over a dozen studies that calculated soda consumers' PED. Soda demand is moderately price- inelastic (PED < 1.0), with a pooled (composite average) elasticity of -0.80, according to one analysis that reviewed many of these econometric studies. In non-economist language, this means an 11% increase in the price of soda would effect, on average, a 9% reduction in the quantity of soda being purchased by consumers.

Because soda's PES is likely to be greater than its PED, it is virtually certain that Berkeley consumers of SSBs would be paying more of the proposed tax than distributors. This expected result is contrary to the Berkeley city attorney's statement accompanying the proposition: "The tax would be payable by the distributor, not the customer." The city attorney's statement is a legal statement based on the proposition's language. It has nothing to do with market realities, nor is it based on the economics of the soda/SSB market. Distributors subject to the tax will do what they can to pass along the tax to retail stores, who in turn will do what they can to pass the tax to consumers. Given SSBs relatively price-inelastic demand, Berkeley customers of SSB-taxed products will pay a higher price for soda, energy drinks and other taxed beverages. Without retail SSB prices increasing due to the tax, consumption of these drinks will not change.

2) How much will consumption of taxed drinks be reduced in Berkeley?  This question, like the first one, depends on what the SSBs' PED is. Unlike some other cities' proposed soda taxes, Berkeley's proposition appropriately names several substitutes that also will be taxed (energy drinks and presweetened teas), and several that will not be taxed (diet drinks, milk products, 100% juice drinks). Are drinks like chocolate milk really substitutes? Maybe, but probably not too close. Exempting milk products likely had much more to do with perceptions of healthful drinks rather than their actual sugar content.

As mentioned above, the more price-inelastic the demand of a good is (e.g., sugared drinks), the less the quantity demanded will change and the more tax revenue will be generated as its retail price rises. The substitution effect is the key driver of these conclusions. If the substitution effect is relatively large and Berkeley consumers don't buy as much 7-UP or other SSBs because of the tax but instead buy chocolate milk or apple juice or other drinks that aren't taxed, then the consumption of taxed drinks will be reduced more than expected (as many proponents of the proposition hope), but revenues accruing from the tax will be reduced (something that proponents may not be pleased with since this will reduce the possible funding for health-related programs). Using the PED value from above and the tax-induced price increase of at most 11% for consumers, the purchases of taxed SSDs in Berkeley might be reduced by at most 9%. Higher soda tax rates could produce greater decreases in SSB consumption, but such tax levels are probably not politically feasible.

It is worth remembering that SSB purchasers in the city can avoid the tax completely either by buying at "very small retailers"[3] in Berkeley (who are exempt from the tax) and/or buying them outside Berkeley's city limits – for example in surrounding Oakland, Albany or El Cerrito. These geographic challenges typify the large difficulties of small areas (Berkeley is less than 18 square miles) in creating effective policies that their neighboring communities do not endorse.

3) Can soda taxes reduce obesity and diabetes?  This last question is more than an economic question; it asks whether Prop D can achieve its underlying, important and needed public-health objective – to reduce obesity and the prevalence of Type 2 diabetes. Unfortunately, available information and evidence for using soda taxes as a fiscal means of reducing America’s growing obesity epidemic is fairly dispiriting. For one reason, no matter what intervention mechanism is used be it a tax or something else, changing people's ingrained behavior is a challenge and can take a long time.[4] The Prop D soda tax is scheduled to end in 11 years.

The Prop D proponents directly link this proposed soda tax to the extant federal and state taxes on tobacco, as a comparable tax to discourage "bad" behavior. Regardless of whether drinking Coke can be rightfully compared to smoking a cigarette (something that seems a stretch to me); the Prop D tax is far lower than the current tobacco tax. The federal cigarette tax is $1.01/pack; the California tobacco tax is now 28.95%, which averages to about 87/pack. Thus, for California smokers the tobacco tax is about 38% of a pack's pre-tax price. Berkeley's proposed soda tax is less than one-third the tobacco tax. Is this sufficient to reduce obesity? It's impossible to say at this point. But the Prop D tax rate is politically expedient because citizens are highly unlikely to vote for a 35%+ tax rate on SSBs.[5] Speaking of tobacco, the Prop D proponents probably don't want the fact known that the 46 states that signed the 1998 Tobacco Agreement, which provided for a minimum of $208 billion in payments through 2023, will spend just 1.9% of their settlement payments and tobacco taxes on prevention programs. This doesn't engender much confidence in how politicians actually spend assigned money.

Another issue in using a soda tax to reduce obesity is that people who reduce drinking SSBs as a result of the tax don't usually switch to drinking water; they mostly consume other drinks that also may have many calories. The effectiveness of a soda tax on obesity would rest in large part on what the net caloric reduction is between previously-consumed soda and the new substitute drink.

Also, there are many factors – like lifestyle – that contribute to obesity and diabetes beyond SSBs. These particulars are not mentioned by Prop D's supporters, who spotlight "Big Soda" as the cause of obesity. And, although these facts shouldn't by themselves diminish the rationale for taxing SSBs, they will make achieving the tax's health objectives, which at a minimum require a net reduction in one's consumed calories as mentioned above, quite difficult. Appreciably reducing obesity will require a sustained, substantive, multi-faceted series of actions, but not limited just to SSB taxes.

To its credit, Prop D would also tax sugared drinks beyond soda that soda aficionados might substitute for if only soda were to be taxed. But the proposition won't tax other high-calorie, high-sugar substitutes like milk (including chocolate milk which is more nutritious, but on a per-serving basis has more calories, far more fat and almost as much sugar as soda and is more expensive) and 100% juices (also more nutritious, and also high-calorie and high-sugar).

An interesting consequence of taxing substitutes beyond just soda, is that prior research has indicated if you design the tax to include a broader number of potential soda substitutes, you are more likely to make people just stick with their original SSB, since they won't be able to "escape" the tax. This will cause SSB consumption not to fall as much and thus raise tax revenues, but not achieve hoped-for health benefits. Ironically, higher tax revenue may allow the Berkeley City Council to fund more obesity-focused health programs, but these revenues are the result of folks not decreasing their consumption of taxed SSBs as much as hoped for. There is also evidence from some studies that have examined how soda taxes have affected consumers, which found that the effects on weight loss were quite modest.

We'll soon see if this notably progressive city's citizens will vote to raise the price of many of the sugared drinks they consume in the name of potentially-improved public health.

A Soda Tax Postscript: Who is paying for the tax?
Summary: After comparing pre-soda tax and post-soda tax prices for Coke sold in Berkeley, I found in January that distributors and retailers are passing along 100% of the tax to consumers. 
I visited a Safeway store in Berkeley on January 2, 2015 to see what prices were being charged for Coca-Cola products sold in the store. As an empirical economist, I wanted to see how the prices of Coca-Cola have changed after the imposition of the now-renowned Berkeley soda tax, passed by citizens on November 4 and enacted on New Year's Day.
I compared these Jan. 2 prices with the pre-tax ones I gathered on Dec. 30 from the same store. In this way I calculated what percentage of the tax is now being charged to and paid by consumers of these sodas and how much is paid by the distributors and/or retailers (like Safeway). Economists call this percentage the "tax incidence".
My results of the tax incidence for several sizes of Coke products sold at Safeway are shown below. This analysis isn't sophisticated or "representative" of all Coke prices now being charged in Berkeley; it simply shows how this Safeway store is pricing many of its Coke products 2 days after the tax went into effect. Nevertheless, it is insightful and offers an unsurprising result, given the dynamics of the retail food market.
Coca-Cola Item
Soda Tax Incidence for Consumers
6-Pack (@16.9oz)
100%
8-Pack (@12oz)
100%
1.25 L bottle (42.2oz)
100%
1 20oz bottle
100%
The price increases I found on Jan. 2 were exactly equal to the tax. Consumers are paying 100% of the tax; the entire price increase induced by the tax. This is diametrically different to what the City Attorney's office said in the proposition statement, "The tax would be payable by the distributor, not the customer." Although the statement is probably legally correct, consumers are paying for the tax. Welcome to the marketplace.
So it seems the Coca-Cola Bottling Co of California - the distributor of Coke products in the East Bay - and Safeway have decided that the demand for Coke and other sugared-sodas they provide to soda drinkers is so inelastic (price-insensitive) that they can simply pass along ALL the tax they are now paying to the City of Berkeley. Perhaps they're right and the marginal impact of the tax is relatively small in terms of changing consumption levels. From a public health perspective, this result may be fine, since if the distributors end up paying for most or all of the tax, then retail prices of sugared-soft drinks wouldn't change much, and neither would consumption patterns.
This result may also be good news for the folks who are counting on soda tax revenues to be distributed to worthy causes (but not before July!); perhaps even to programs that can reduce youth obesity, if the City Council allows it. In the next few months, we hopefully will learn how sales are affected. All I determined here was the price change and tax incidence, not how consumer sales are changing because of the tax-induced price increase. How sales actually change with these increases will establish how much revenue is available to the city, and ultimately whether the health of Berkeley consumers of sugared-soda may be improved. Time will tell.
Stop the presses. Another assessment of the effects of the soda tax was published in August that offers a much different result than what I show above. This analysis apparently gathered price data from all Berkeley’s groceries, supermarkets, pharmacies, convenience stores and gas stations. Prices in these stores were then compared to a sample of shops in San Francisco. Why San Francisco stores were selected doesn’t make much sense, San Francisco’s economic marketplace is very distinct from Berkeley’s. The study should have gathered information from Oakland or El Cerrito stores to offer a more appropriate apples-to-apples (or is it Coke-to-Coke) comparison.
Nevertheless, this assessment found that just 21.7% of the Berkeley soda tax was passed along to customers. The Daily Caller stated that Berkeley’s store owners have refused to play ball and have only passed on a fraction of the price increase to consumers. “These results imply that the Berkeley soda tax, because it is passed through to consumers to a lesser extent than anticipated, will result in less of a reduction in consumption, and thus less health improvement, than anticipated,” the study said.
The Daily Caller's characterization makes it seem that it’s the store owners’ public responsibility to raise the price of sugared-soda the entire amount of the tax. The stores here being blamed for only raising the price 21.7%, not 100% (as I found). But if this study is correct, then the City of Berkeley will have a lot less funds to spend addressing the city's perceived public health care needs.



[1] The reason the Berkeley proposition doesn't state its tax revenues will directly fund specific health programs in the city is that such specific tax propositions require a two-thirds yes vote; general-fund taxes like Prop D require only a simple majority. Initial surveys indicated that less than a two-thirds majority of Berkeley voters would likely support such a tax.
[2] Incidence of a tax is a measure of which market participant(s) – consumers and/or producers/sellers – actually pay the tax.
[3] Very small retailers, according to Prop D, are those that have annual gross revenues of less than $100,000. As defined, these Berkeley retailers are indeed very small, with total daily sales of less than $300 if they are open 7-days a week. A typical 7-11 store probably grosses at least 10 times that much each day.
[4] Changing habitual personal behavior – even if it's unhealthy – can take significant time and effort and require multiple methods. It's been 50 years since the first US Surgeon General's report on smoking was published. This report directly linked smoking to disease and shorter life spans. Substantial taxes have been imposed on tobacco products as well as wide-spread, multi-media anti-smoking advertising-public service campaigns. With these considerable activities, smoking rates for adult Americans have declined significantly over 5 decades. And the latest CDC report states that despite these efforts 42.1M Americans remain current smokers, representing 18.1% of our adult population.
[5] The proposed San Francisco soda tax is 2₵ per fluid ounce.

No comments:

Post a Comment