Do Sin Taxes Actually Work? | Ben Lockwood

Studying the Effects of Philadelphia’s Soda Tax

Dan Loney: Have you ever heard of a sin tax? It’s a type of tax put forward by a city or a state, primarily to add a cost to something that may not be the healthiest or the best for us. There have been sin taxes on things like soda, tobacco products, alcohol and others, and they may be growing in use to raise needed revenues. But do sin taxes actually impact the public’s thought process enough to change their behavior? This is research conducted by Wharton’s Ben Lockwood, who’s an assistant professor of business, economics and public policy, and he joins me here in the studio.

Great to see you, Ben. Thanks for your time. The research looks into the mindset of people as these sin taxes are implemented, and whether or not they actually have an impact on their purchasing habits?

Ben Lockwood: That’s the idea. I’ve done a couple of research papers in this area, ranging from the general theory and an understanding of how sin taxes work and how effective they might be, when you might want to use them. Obviously, they’re used on a bunch of different products, from alcohol to cigarettes to other forms of tobacco, to now sometimes gambling or soda. More specifically, I’ve zeroed in on a couple of particular types of sin taxes. One big set of projects on soda taxes, another set of projects on gambling in the form of state-run lotteries, to kind of understand what the trade-offs were and what the behavioral effects were in those specific areas.

Loney: Here in Philadelphia, a sin tax on soda was put into place. What kind of research do we know from having that and whether or not it had an impact on sales here in Philadelphia?

Lockwood: It’s been maybe close to a decade ago [that] Philadelphia implemented a 1.5 cents per ounce tax on sweetened beverages. Soda, diet soda, and some other kinds of non-carbonated, sweetened drinks. And 1.5 cents per ounce might not sound like a very large amount, but the average cost of soda in the U.S. is something like 4 to 5 cents per ounce. So, something like a 30% tax. It’s actually pretty substantial.

There’s enough of a track record, enough evidence over these last several years to see what sort of effects that’s had. My read of the current evidence is that that tax reduced total soda purchases in the city of Philly by close to 50%. Around half of that reduction reappeared outside the city as cross-border purchases. You go into Jersey, you go up to the Main Line or wherever those taxes aren’t in effect, and you buy your soda there instead. Some of it gets offset by that kind of cross-border shopping. But still, that’s something like a 25% reduction in total purchases, so it’s not trivial.

Loney: It’s interesting that people would consider not buying their soda within the city limits and actually go outside the borders. I guess it’s probably people who are close to the borders in general. Go a mile, go 2 miles to buy their soda and then bring it back into their residence in the city limits.

Lockwood: I think that’s right. It’s people who sort of live close to the edges. I think there’s some evidence for being geographically close to the edges, and also being people who kind of chain their trips together and do a large set of purchases at the Costco nearby or something, and that’s when you get all your soda. In the research that we do, one implication is, if you do want to do this sort of sin taxation, it can be beneficial to try to implement those taxes on a larger geographic area. Because the smaller the area that these taxes are confined in, the more you lose any of their revenue-raising ability and any of their desired behavioral effects to people who are just cross-border shopping.

Loney: It’s interesting you said that the amount of soda purchasing went down as much as it did. You wonder whether or not there’s actual recognition by the public of the difference in cost because of that sin tax component.

Lockwood: This is not specific to sin taxes. Often, in economics generally, people are surprised to find out that small price changes can nevertheless generate a change in total behavior. The price of a $30,000 car — or these days, it’s probably a $60,000 car — goes up by $1,000, and it’s like, does anyone pay attention to that? But you can see in the data, demand curves slope down, just like we teach in our MBA classes. You can see in the data that there were some people who were a little bit on the margin between this and something else, or between buying now or waiting till next year, and you do see some effect in response to that. This is sort of echoing those general results, that even small price changes do create changes in how much people purchase.

How Sin Taxes Affect Consumers Long-term

Loney: Can you determine the larger-scale impact on health and maybe other components because of a lowering in the consumption, or at least the purchase, of soda within the city of Philadelphia?

Lockwood: You might be able to see this sort of thing eventually. From the way that we think that health consequences of sugar consumption operate, you might not actually expect these things to materialize for 10 years down the road or something, with reductions in diabetes contraction. Even then, many people who lived in Philadelphia at the time might now live in Florida or somewhere else. So, I would be surprised if we’re able to pick up very much direct effect for this quasi experiment of the soda tax being implemented here. But there is evidence more generally of the effects of sugar consumption on health down the road, and other types of consumption. Certainly, in the case of cigarettes, for example, there’s a lot of evidence about the linkage between smoking and cancer.

Loney: It’s probably not a surprise that the types of products that jurisdictions may put a sin tax on are growing in the last few years as well.

Lockwood: I think that’s true, partly because there is this growing realization that some products might have these health consequences. From an economics perspective, the way I think about this is not necessarily that you need to tax things just because they have harmful health consequences. There are lots of things that might be harmful for one’s health, but people are aware of those harms, and they’re interested in doing the thing. We don’t necessarily tax mountain biking or rock climbing just because they can be dangerous.

If there is some aspect of a cost that people aren’t thinking about or aren’t taking into account, either because of behavioral biases or because they aren’t aware of the information, then what you want to try to understand is, if they were fully informed and really understood the context and consequences of all their actions, would they be consuming less than they currently are? If so, that’s a situation where you might want to impose a tax to bring their actual consumption into line with what they might rationally do if they took all those costs into account. It’s pretty much the same way that we have this rationale for imposing a gas tax so that people take into account the negative externalities, the pollution consequences, or the congestion consequences of buying gas, and bring it into line.

Loney: But from a public policy perspective, these have become important revenue drivers for cities and states.

Lockwood: It’s true. This is one of the ironies of sin taxation generally, is I think they’re sometimes looked to as a potential source of revenues. But if part of the goal of the tax is to reduce people’s consumption of something that might be harmful for them, then if the tax works well and people reduce their consumption a lot, it necessarily doesn’t bring in as much in revenue, right?

The way I think about the revenue question is the taxes that you need to bring in are determined by the jurisdiction’s spending decisions and their priorities. You’re going to have to cover that spending one way or another. If you don’t raise it via a soda tax, you’re going to have to raise the sales tax or the personal income tax or something else. It’s a question of which different things are we going to tax, rather than should we tax soda at all?

Key Considerations for Tax Policymakers

Loney: What do you think the larger discussion is about the impact that sin taxes are having in communities?

Lockwood: I think there are a couple pieces to this. One is the kind of behavioral consequence piece. Are people changing their behavior? Is that making them healthier or not? To what extent is it having an effect? Another piece of it is, are there distributional consequences to this? Which is the fancy way that economists talk about being really regressive or progressive with your tax choices. Are you going to tax poorer people much more heavily, or richer people more heavily?

One thing that we know about lots of these goods — soda, cigarettes, to some extent alcohol — is that they’re more heavily consumed by people who are sort of lower in the income distribution. A thing you want to be aware of, a thing you want to be careful of, is are you increasing taxes that are going to fall most heavily on people who are poor? Now, there are ways of offsetting that with making the income tax a little more progressive at the same time that you impose some sin taxes in a way that offsets that distributional impact. But it is something that you want to have in mind when you’re deciding on these kinds of tax policies.

Loney: What’s your expectation for looking at this research and taking it forward?

Lockwood: It’s a good question. There are a number of increasing future areas of potential sin taxation that I think are becoming relevant. There’s been an explosion in various forms of online gaming, sports betting, this kind of stuff. There’s also increasing marijuana legalization, and that’s viewed as a potential sin tax base.

With those sorts of areas, my question would be, to what extent are people buying those things or doing those activities more than they would if they were fully rational or fully aware of the costs or fully cognizant of this stuff? A flip side of that is that I think that, for political reasons, there’s sometimes a drive to tax these new markets just because they don’t have an entrenched body of stakeholders already that are existing and that can oppose a tax. Now, from an economist point of view, that’s not really a good reason to tax them, just that nobody is currently consuming them a whole lot. You want to think about whether there are these behavioral biases, and what the revenue trade-offs are.

Loney: Are there sin taxes in a majority of states or a lot of communities right now?

Lockwood: Yeah. If you look at the set of different sin taxes that are available, lots of states have alcohol excise taxes, cigarette taxes. Many states have state-run lotteries, which I do think of as a taxed good, in the sense that a portion of the revenues are withheld before the prizes are paid back out, and that goes into the state’s coffers. Between those things, you’re at a point where most states and increasing number of localities, of cities and smaller areas, are imposing these kinds of taxes, too.