I think an excellent Pigouvian tax would be a tax on emails. Many emails involve a negative externality (I don’t really want to receive them) and almost all the ones I really want to get are worth much more than a penny or so to the sender. So a penny tax (say) on email would probably generate large amounts of revenue, mitigate an important negative externality, and have minimal inefficient disincentives. Since email servers are necessarily centralized and networked and all email senders are ipso facto connected to an ISP who is charging them for access the transactions costs and evasion problems seem low.

As a reminder, a Pigovian tax is one that is levied to change market behavior – typically to address a market failure like externalities or to prevent over consumption of a common good.  Mankiw extends the idea, hoping that the recipient could set the price. If someone wanted to seriously reduce the email they received – limiting to only those messages from people who really care – then they could set a high price for allowing a message to appear in their inbox.

Interesting idea. It’ll never happen, but it is interesting nonetheless.

Pizza Hut Double Deep Pizza

I love the cheese, the crust, and most of the other, assorted things that can go on a pizza. Sadly, many of those things are not good for me. This article, put out by Reuters, reports on a study measuring the impact of a tax on pizza on the number of pizza and soda calories consumed.

Over a 20-year period, a 10 percent increase in cost was linked with a 7 percent decrease in the amount of calories consumed from soda and a 12 percent decrease in calories consumed from pizza.

The team estimates that an 18 percent tax on these foods could cut daily intake by 56 calories per person, resulting in a weight loss of 5 pounds (2 kg) per person per year.

OK, budding economists. What does the first sentence sound like? If you said price elasticity, you would be right. They measured the change in demand of a product (with calories as the proxy for quantity of pizza) as a result of a change in price. Is calorie demand (at least for pizzas) price elastic or price inelastic?

Then, to add a policy bent to it, consider the idea of Pigovian taxes. If obesity is a negative externality (and it is – since obesity leads to higher health care costs and to higher insurance premiums, even among less heavy people) then one could estimate the impact of a pizza tax on consumption of those oh-so-good, but oh-so-bad for you consumer food items.

Here’s a quick list of reasons to tax. Each have a bit longer explanation down below.

We tax to…

1. pay for public services that are easier or more efficient to provide as a community than to pay for individually.  (AKA public goods)
2. correct for inequalities in individual wealth or income – to provide some basic level of food, shelter, medial care, etc. (AKA welfare and other social services)
3. correct for externalities.
4. Change behavior – encourage or discourage through incentives

1. Public Services/Public Goods:  There are services that many people want, but individually we could not afford to buy them. While it is possible for groups of individuals to come together privately to pool their funds and provide the service, they often run into the free rider problem. So we give government the ability to build roads, provide police and fire protection, and many other worthwhile goods and services. Often times voters have to approve the tax to pay for these. Here are some other posts on the topic.

2. Income Redistribution / Social Services: In some economies there is an explicit goal for a Robin Hood policy (take from the rich and give to the poor) – purely to even out income or wealth. In the United States and many other countries there is a social ethic or value that says that the poorest members of society should be able to live in at least some minimum level. This ethic or value is controversial, of course. Some voters support strong government efforts to improve the lives of our poorer members – along the lines of European social democracies. Other voters prefer a self-determination, self-reliance model, where citizens have opportunities but are left to their own to survive or advance in the world. And many voters are somewhere in between. In a later post we’ll show how many tax strategies focus on extracting more tax revenue from the wealthy than from the poor.

3. Correcting Externalities: In economics we call the presence of externalities a market failure. When someone outside of a market transaction is either harmed by or benefits from that transaction, we have an externality. If there is an externality then the market will produce either too much or too little of that good, and will not reach a social optimum. The most common example is factory pollution. When the factory produces a good and pollutes while doing that, others outside the factory are affected. See more on externalities here.

4. Incentives to Change Behavior: Some times we enact taxes to discourage behavior. “Sin taxes” are one such example. Tobacco taxes can be used to defray state health care costs. Generally this is a pretty inefficient mechanism since many of the “sins” are addictive and demand for them price inelastic. More often a tax on tobacco or alcohol is just a convenient way to raise revenue, and voters are more likely to support those taxes over more general income or sales taxes. A more modern application of this category is called a Pigovian tax, and the prime example being discussed is a carbon tax. Taxing activities and products that release carbon into the atmosphere should discourage and reduce those activities.  We also give tax credits for behavior we want to encourage.

Here’s an excerpt:

In the years leading up to his death, in 1959, he was a reclusive figure, rarely venturing from his rooms at King’s College. His novel ideas on taxing polluters and making health insurance compulsory were met with indifference: Keynesianism was all the rage.

We’ve talked about Pigou, and his Pigovian taxes in other posts. As the article reminds us Pigou recognized that sometimes market dynamics fail to address social needs. One set of these market failures (or imperfections as they are also called) is externalities. These occur when players that are not a part of a transaction are affected by the transaction. Neighbors living next to a polluting plant do not participate in the sale of the goods or power produced by the plant, yet they incur a cost. The friends of a child who has been immunized benefit from that child’s vaccination, yet they don’t get involved in the cost-benefit decision made by the child’s parents. Both are cases of externalities. A Pigovian tax can help remedy a negative externality, providing incentive to the polluter to change their behavior.

Slate published this article, “Sweet Surrender: Taxing soda to make you stop drinking it.” This is another example of using an excise tax to change consumer behavior. If one believes that there are negative externalities to the consumption of sugar-laden, artificially-flavored, carbonated water, then these taxes fall under the Pigovian Tax label.

Thanks to a tip from Greg Mankiw’s blog – we learn that Oregon was the first state to impose a tax on gasoline – back in 1919. This article in Wired describes the interest of the state to fund highway projects, including the Columbia River Highway up the Gorge. So, this isn’t really an example of a Pigovian Tax, which is usually imposed to change behavior.

What do you imagine the price elasticity of demand for gasoline was back in 1919?

Here’s a recent example. From the Chicago Tribune on November 14, 2008:

Want to stop people from drinking too much? Forget earnest public service announcements. Just make alcohol more expensive.

A study released Thursday by the American Journal of Public Health found that when the tax rate on alcohol went up, deaths caused by drinking went down.

So, in this case the purchase and consumption of alcohol has a negative externality – a larger number of deaths in society. A higher death rate imposes a social cost – to innocent victims of drunk drivers, to lost productivity, and to related damage. A tax on alcohol is a Pigovian tax. It does not ban the purchase of alcohol, but the higher resulting price provides an incentive to change purchase and consumption behavior. The study mentioned in the article provides a link between the tax and reduced alcohol-related deaths.

The principal challenge in creating a Pigovian tax is estimating what the social costs are, and setting a tax that recovers those costs. And, we know that imposing a tax, particularly an excise tax on something like alcohol, results in deadweight loss – a net loss in combined consumer and producer surplus.

“…come again?…” Here’s what I mean…. Let’s assume that a coal burning electric power plant produces pollution as a by-product of power generation. The pollution imposes costs on the neighbors – impacting quality of life, or health issues, property values, etc. So the pollution represents a negative externality. Instead of using direct government regulation, where laws and regs require certain actions, a different option is to add a tax that matches the external costs being created. If there are $50 million in external costs a tax of $50 million is imposed. Ideally this tax goes up and down with the amount of pollution released. Here’s the Wikipedia write-up.

If done correctly this Pigovian tax will change the behavior of the producer – the electric power plant. They will adjust their output (usually lower) based on the various theories on how producers choose optimum output. The important point is that those external costs are now being considered, rather than ignored.

Two important points about a Pigovian tax. First, there is a direct financial impact on the producer, but they are left with the option on how to minimize the tax. They can choose smoke stack scrubbers, or they can look for a better coal, or they can reduce output. The choice is theirs and in theory they will make a more efficient choice than a standard solution required by the government. The other important point is that funds are collected by the tax and can be used to reimburse those harmed by the pollution.

There is one remaining downside to a Pigovian tax. Like many such taxes, there is deadweight loss, where the combined consumer surplus and producer surplus show a net loss after the imposition of a tax. Here we have to weigh the relative importance of the deadweight loss against the importance of trying to solve the negative externality.

Greg Mankiw (currently teaching at Harvard and previously chair of the President’s Council of Economic Advisors under Bush) is a fan of the Pigou solution and has formed an informal Pigou Club. See this post for a recent example, from a candidate for the U.S. Senate in New Hampshire.