On Feb. 22, Berkeleyside reported the results of a UC Berkeley study. The headline read, “Sugary drink consumption in Berkeley down more than 50% since introduction of soda tax.”
There are two problems with this headline. First, it exaggerates the results of the study, which, as its authors pointed out, was confined to a particular subset of Berkeley’s population. Second, the results of the study are contradicted by the history of the soda tax revenue.
State legislators in Sacramento are considering several bills regulating sodas and other sugar-sweetened beverages. For the sake of crafting thoughtful policies, our legislators need accurate information about the effects of existing soda taxes. In this op/ed, I attempt to provide some of that information.
The UC Berkeley study was based on self-reported results that relied on the memory of the surveyed subjects. The study did not interview the same residents over time, nor was it based on any diaries of food and drink consumption.
There is an alternative set of data that, while still imperfect, are far more objective — the history of soda tax revenue in Berkeley. This data tells a very different story. Here is a graph of the Berkeley soda tax revenue:
A quirky feature of Berkeley’s soda tax is that it should be self-defeating. If it leads to lower soda consumption, the tax revenue should decline. So far in Berkeley there is no evidence that is happening. Soda tax revenue, which reflects soda consumption, fluctuates month-to-month, but does not have an obvious downward trend.
However, there are several confounding factors that could complicate these numbers. First, soda sales nationally have been declining for years. This could lead to a downward trend Berkeley’s soda tax revenues that has nothing to do with Berkeley’s soda tax. Over time, Berkeley residents could shift their soda purchases to Costco outlets or other stores that don’t charge the soda tax.
On the other hand, information provided to retailers about the tax and better enforcement of it could lead to a higher collection rate, causing the trend to rise over time. And of course, a mix of these factors operating together could partially obscure the underlying trends.
To delve deeper into the data, we can enhance the graph in two ways. First we can add data from the city of Albany, just north of Berkeley, which adopted a soda tax modeled on Berkeley’s in Nov. 2016. Second, we can switch to a logarithmic scale so that each increment of the vertical axis represents a doubling of soda tax revenue.
In the new graph, if Albany’s soda tax revenue falls by 50 percent from $32,000 to $16,000, that creates the same vertical drop in the graphed data as a 50 percent decline in Berkeley’s revenue from $128,000 to $64,000. This makes it easy to compare percentage changes between the two cities.
Berkeley has had a few spikes where the revenue either doubled or fell by almost half in some months. But on average, it tracks around $137,000 per month. Albany’s revenue appears to start high and decline, but that is due to delayed payments at the start of the program that covered more than one month. Albany’s soda tax revenue has averaged about $22,700 monthly, and doesn’t show any long-term declining trend.
Berkeley (population 122,000) and Albany (population 20,000) are similar in another way. In both cities the revenue is about $1.13 per person per month. Given that the soda tax is one cent per ounce of soda, a quick calculation shows that the average consumption is 26 oz. per week, a little more than two 12 oz. cans of soda per person per week. That isn’t a public health crisis. However, averages can be deceiving. It could be that many people drink less than average, while a few people drink far more.
There is another problem with the soda taxes — the taxes raise the wholesale cost of sodas, but the cities depend on retailers to pass these cost increases on to their customers. Yet we know that this doesn’t always happen. Two studies of the Berkeley soda tax show that only about half of the tax is passed on. In Albany, based on my survey of fast food, grocery and supermarket retail stores, the pass through is close to zero. I could find only one store in Albany, the Solano Avenue Safeway, that passed on the tax at all; and then only incompletely.
If retail outlets are not passing on the tax via the retail price of sugar-sweetened sodas and other beverages, then the stores are either absorbing the loss (which is unlikely due to the narrow profit margins in grocery retailing), or raising the cost of other items in their stories. To a large extent, the soda tax is actually a tax on food.
As for the expenditure side, if the soda tax funds new health-related expenditures that otherwise would not have been made, then the tax is fulfilling its goals. However, if the pot of soda tax money is being raided to fund existing health programs, then the net result of the tax is to free up money in the city’s general fund for other purposes.
Perversely enough, the Berkeley and Albany soda taxes are a win-win-win. Cities get more revenue, voters get to feel satisfied they did the right thing, and the soda companies are delighted to learn that soda taxes barely affected their sales. But the reality is that the soda taxes don’t seem to work — at least not yet.
I believe we have the obligation to let our legislators in Sacramento know just how well our soda taxes are working. With this information, they can craft more effective policies to reduce the human and social costs of poor nutrition in California. The cities of Berkeley and Albany should also reconsider what they are able to accomplish with their soda taxes.