The Environmental Case for a Progressive Consumption Tax

And Why it is Preferable to Non-Linear Pricing

My colleague and co-author Alexandre Chirat published with Basile Clerc an op-ed in Le Monde a few days ago defending the use of non-linear pricing as an efficient and fair way to reallocate resources in the fight against climate change (in French). They focus on the case of electricity consumption:[1]

“Let's take the example of electricity, part of whose production will have to respond to the decline in combustion-powered cars. Non-linear pricing means charging a household a price per kilowatt-hour that varies according to consumption. Let's imagine that for a couple with two children in an energy class D home, consumption is defined as “normal” at 12,000 kilowatt-hours per year. The first 12,000 kilowatt-hours consumed are billed at 0.21 euro (off-peak rate). But between 12,000 and 14,000 kilowatt-hours, the price rises to 0.27 euros (peak rate), and then increases exponentially. Such a measure encourages all households to save energy, while also having a social impact.”

Non-linear pricing is an attractive alternative to direct price control methods such as price- ceilings/floors or restrictions on quantities. It preserves the information conveyed by price and implements a redistribution from wealthy households (who presumably have high energy consumption) to poor households. The tax proceeds could also fund investments in the energetic transition toward less polluting technologies. Alexandre and Basile also suggest that the same mechanism could be used to affect consumption choices at the margin regarding, for instance, transportation means. Instead of strictly capping the yearly (or even life-long) number of flights that one individual could have, a non-linear pricing scheme on plane tickets would potentially have a strong substitution effect. As the authors note, they are by no means the first to propose such a mechanism – the Nobel Prize Joseph Stiglitz already made the same suggestion two years ago for the electricity market.

Regular readers know I’m not a friend of price controls at all. Non-linear pricing is, however, attractive because it behaves more like a progressive tax on some activities. Of course, there is an element of administrative control insofar as there is a somehow arbitrary scheme that “artificially” increases the price with the level of consumption. As always when there is an interference with market prices, we must ask what counts as a “normal” or “desirable” level of consumption. Compared to a regular Pigouvian tax, there are in particular two difficulties. First, part of the justification of non-linear pricing is unrelated to externalities per se. The justification is mainly to push for a reallocation of resources as part of a long-term industrial policy aiming at the emergence of new technologies. While you can eventually try to evaluate the social cost of an externality to set the tax at the correct level, such an evaluation is harder to realize here. This introduces a lot of arbitrariness in the scheme, and eventually a risk of “regulation capture.” Second, non-linear pricing, at least in the proposal of my two colleagues, ostensibly mixes efficiency and distributive purposes. This is in general not a good idea. As the economist John Cochrane puts it, “don’t distort prices to transfer income.”

Now, I nonetheless think that there is something worth in this proposal. The point however is that there is a far simpler, less bureaucratic, and therefore less prone to “regulation capture” method to achieve similar effects. This method is the implementation of a steeply progressive consumption tax. The idea of a progressive consumption tax was endorsed long ago by Milton Friedman as the best way to pay for the war effort. More recently it has been essentially defended by the economist Robert H. Frank.[2] Frank’s defense of a progressive consumption tax takes place within his discussion of the social cost of the consumption of positional goods. As I explained here recently, in a different context, positional goods tend to generate a “non-material” kind of negative externality when they are consumed because, while they improve the consumer’s social position, they in the meantime negatively affect all other persons who care about their relative social position. On top of this negative externality, for some categories of the population at least, an increase in the demand for positional goods goes hand-in-hand with an increased willingness to pay for those goods. The result is a collectively harmful arms race.

Frank argues that the best way to deal with the social cost of the consumption of positional goods is to tax it. An obvious problem is that any classification of what counts as a positional good is likely to be controversial. After all, there are many goods that some persons may consume for positional reasons while others don’t. This is fairly subjective and in general, we don’t want to leave this appreciation to a political authority, even a democratic one. Hence, the best way to circumvent this problem is simply to tax all consumption activities by introducing a steeply progressive tax scheme – this differentiates the progressive consumption tax from the standard VAT. The rationale is that at the highest marginal tax rates, households will arbitrate against essentially positional consumption goods. In practice, that will barely affect their well-being, while halting the positional arms race. How would that work exactly? Frank suggests that over the long run the progressive consumption tax would replace the current income tax:

Replacing the current income tax with a progressive consumption tax is the only way to cover our current revenue shortfall without demanding painful sacrifices from voters. Such a tax, which has been proposed both by conservative economists like Milton Friedman and liberal economists like Edward Gramlich, would be simple to implement. Families would report their incomes and their annual savings to the IRS, just as many now do with 401(k) and other similar retirement savings accounts. Their taxable consumption would then be calculated as income minus savings minus a large standard deduction–say, $30,000 for a family of four. For example, a family that earned $50,000 and saved $5,000 during a given tax year would have taxable consumption of $50,000–$5,000–$30,000, or $15,000 total. Tax rates on taxable consumption would start off low–say, 10 percent for the first $30,000 of taxable consumption. Under the consumption tax, this family would owe $1,500, about half of what it would pay under the current income tax.”

Note how this works similarly to the non-linear pricing scheme, except for the fact that it does not discriminate which consumption activities are concerned. Compared to current progressive income taxes, a progressive consumption tax has no disincentive effect on work – or at least a less direct one. It is in this sense closer to an optimal tax. Moreover, as it encourages redirecting the use of resources from consumption to investment, it would contribute to improving the productive capacities of the economy, and eventually (combined with other inventive schemes) accelerate the development of carbon-free technologies. Note also that, at least in the short term, a progressive consumption tax could coexist with an income tax.[3] The most important point is the progressiveness. We may want to exempt households whose consumption is below some level, eventually fairly high, as in Frank’s example above. But then, marginal rates should increase fast.

Of course, many details would need to be filled in to assess this proposal. However, compared to a non-linear pricing scheme that applies only to a small set of consumption activities, I only see advantages. The main one is that it considerably lessens the risk of regulation capture. Since the tax applies to all consumption activities, there is no need for intricate negotiations about which activities should be non-linearly priced and how. Moreover, as there is a strong correlation between the richest households, the absolute level of consumption, and the level of pollution emission, a progressive consumption tax would for sure strongly incentivize the wealthiest to reduce the part of their consumption activities that pollute the most, such as traveling by plane several time a year. The over-consumption of positional goods is itself a source of pollution that would be reduced by a progressive consumption tax. And, again, it would reallocate resources from consumption to investment, which is exactly what we need to operate the transition toward less polluting technologies.

---

[1] Translated with the help of DeepL.

[2] Milton Friedman, “The Spendings Tax as a Wartime Fiscal Measure,” The American Economic Review 33, no. 1 (1943): 50–62. Robert H. Frank, The Darwin Economy: Liberty, Competition, and the Common Good (Princeton N.J.: Princeton University Press, 2011).

[3] Laurence S. Seidman, Pouring Liberal Wine into Conservative Bottles: Strategy and Policies (Lanham, Md.: UPA, 2006).