Consumption taxes (VAT, sales tax) are mostly applied at a flat rate. Some countries have exemptions or lower ratings for essential goods. But even with that, the tax is much less progressive than other taxes, like income tax.

There have been some suggestions that countries should move away from income tax and corporation tax and generate most revenue from consumption taxes. I just mention this as background, I know there's no clear consensus, and my question is not about that.

If consumption taxes were used more than income tax, how could these be made progressive?

Edit: I did some searches for progressive consumption tax and found this article that is most promoting a book and presupposes knowledge of terms like Flat Tax. Wikipedia on X Tax is interesting but brief. Further elaboration would be informative.

  • 1
    $\begingroup$ @1muflon1 - Understood, and thank-you. I included some links I found while researching. To be honest, I was mostly hoping for a good answer in the style of Stack Exchange as these are often better than the info you can find on Google. Hope asking for this reason is allowed on this site? I'm a layperson who's not studied economics so apologies if the question demonstrates basic misconceptions. $\endgroup$
    – paj28
    Mar 3, 2021 at 19:07
  • $\begingroup$ Yes, that is fine, and now your question is better so I removed the comment, but as the comment said if you tell us what you already searched it helps us understand the question better. In addition, often you can find excellent explainers by just quick google search and having this policy helps us prevent people using SE as a substitute for search engine $\endgroup$
    – 1muflon1
    Mar 3, 2021 at 21:43

3 Answers 3


Trying to make consumption taxes more progressive is not impossible but it is uphill battle since any consumption tax is inherently regressive since lower income people spend larger share of their income on consumption than higher income people (see discussion of this in Blanchard et al. Macroeconomics: A European Perspective).

Probably the most straightforward way how consumption tax can be made more progressive is to tax goods consumed more frequently by upper income people at higher rate than the goods consumed by lower income people.

While this is doable in a sense that it is technically possible to implement, it is often not very economically efficient for various reason. Here are some of them:

  1. Luxury items tend to have higher price elasticity than necessities. One of the most basic rule of taxation is the famous 'Ramsey rule' that suggests that inelastic goods should be taxed more than elastic goods (see Stiglitz Economics of the Public Sector). Hence, such attempt to make consumption tax more progressive goes against the most basic rule of optimal taxation.

  2. Consumption does not carry information about your type (in terms of income) perfectly. Low income people might buy luxuries even if they do that much less often than high income people, and there are high income people who might distaste luxury consumption (e.g. consider Bill Gates who famously dresses modestly) and thus while doing this can make the tax more progressive it is to a degree tax on a lifestyle rather than a fully being more progressive tax. For example, in the Netherlands where I live food for bunnies and guinea pigs was for long time taxed with different VAT (presumably because government determined that rich people prefer bunnies (bunny food had higher VAT), completely ignoring they are in fact perfect substitutes).

  3. Research shows that if government has access to both consumption and non-linear income taxes (i.e. government levies them at the same time), it is always more efficient, even with Ralwsian social preferences (the most redistributive preferences there are), to have flat consumption tax and highly progressive income tax. This is because income actually caries information about your type. Thus if you care about having progressive tax system, and you consider tax system as whole not just separately, and you want to achieve maximum progressiveness, it is better to do that all through income tax rather than trying to change VAT (see Atkinson and Stiglitz 1980. Lectures on Public Economics or Jacobs Principles of Public Finance).

    In fact, from economic perspective there is little difference between income and consumption since all income is eventually spent on consumption. Saving is just postponed consumption (although saving can have beneficial effect on growth). However, income carries with itself information that can be more effectively used for redistribution than consumption that carriers very little information usable for redistributional purposes. Consequently, if your goal is maximum progressiveness of tax system you would optimally achieve that all through income tax, and the only reason why you would not set consumption taxes to zero is to alleviate some distortions that income taxes cause on labor market (As a matter of fact in the special case when utility function is weakly separable between commodities, and labor the consumption taxes would be fully redundant - this is known in public econ as the Atkinson-Stiglitz Theorem). Using progressive consumption tax is only optimal in situations where you forbid government from using progressive (non-linear) income taxes or income taxes altogether (or perhaps there might be some special cases - but full review of literature on this issue is beyond scope of SE answer).

There are some other proposals such as the X-tax you mentioned. However, note X-tax is not a pure consumption tax. X-tax has both consumption and income component as X-tax also combines consumption tax with taxes on wages and so on (see Bradford, 2008). Consequently, I do not believe these proposals would qualify for what you are looking for in the question since you ask conditionally on "if consumption taxes were used more than income tax".

Next, you could also make consumption tax progressive by applying consumption tax at a flat rate first and afterwards let people with low incomes claim tax rebates on their consumption. This is probably the most efficient way how to make consumption tax more progressive (even though most of the previously mentioned caveats still apply). This tax can be designed in various different ways, you can read more about it in for example Seidman (2013) or Seidman: The USA Tax A Progressive Consumption Tax. However, once again you can't get away with involving those 'pesky' incomes in the tax system design. Again this is because incomes carry most information relevant to progressive taxation, but in this case I think it would satisfy conditions imposed in your question since even though this consumption tax is based on income (since low income people get rebates), it does not necessarily involve directly levying income taxes (such as X-tax, which includes extra taxes on labor earnings).

  • 1
    $\begingroup$ Thanks for the detailed answer, this is exactly what I was looking for and very informative. I will leave you a bounty. $\endgroup$
    – paj28
    Mar 4, 2021 at 11:05
  • $\begingroup$ @1muflon1: You compare luxury good only with necessities, I think that is an unfair comparison. If we compare luxury goods in the sense of the word (and ignoring the difficulties establishing that) with other luxury goods that are not "luxurious" in the sense as most common people understand it, then I would have a strong prior that the former are more price inelastic than normal good. This is one of the reason why I like luxury brand stocks such as LVMH, they have a lot a price power. To put a blunt example, people buying Rolex watches don't really care if the watch costs 2000 or 2500 Euros.. $\endgroup$
    – Papayapap
    Mar 11, 2021 at 15:14
  • 1
    $\begingroup$ @Papayapap 1. luxuries have empirically high elasticities, in fact in literature luxury good is often defined in terms of having high elasticity (usually higher than $|e|>1$, see discussion in Kemp 1998). 2. Price elasticity of demand changes across income. Sure person at a top 0.1% of income distribution may not care much between 2000 or 2500e but person in top 30% 20% or 15% will likely care. Even for Veblen goods where the elasticity is positive it will only be positive for people above certain incomes $\endgroup$
    – 1muflon1
    Mar 11, 2021 at 16:59
  • $\begingroup$ (Some comments have been removed. Please keep comments in the chat on topic.) $\endgroup$
    – Kitsune Cavalry
    Mar 16, 2021 at 3:16

This answer simply points out what is, I think, an important part of 1muflon1's answer. There are two ways to think of consumption taxes: One is to tax certain consumption goods with a VAT. Making this progressive is an uphill battle and will cause a lot of distortions that have little to do with distributional concerns. On the other hand, you can simply tax the money spent on consumption, and this can be done in a progressive way. Income equals consumption plus saving. So by making savings tax-exempt, you turn a progressive income tax into a progressive consumption tax.

  • $\begingroup$ Very interesting take! I know many countries have schemes (like ISAs in the UK) where interest on savings is tax exempt, and also that pensions can be paid into tax free. Are you aware of any countries that have a system like you describe? $\endgroup$
    – paj28
    Mar 4, 2021 at 11:14
  • $\begingroup$ @paj28 No, I just heard the theoretical point being made once. I forgot who made it though, $\endgroup$ Mar 4, 2021 at 11:58
  • 1
    $\begingroup$ +1 I recall that this sort of tax was discussed (but not implemented) in the UK in the 1980's. There was even talk of abolishing income tax and replacing it with such a tax. However, defining consumption in this context raises points which may seem technical but have a major effect both on the amount the tax would raise and how progressive it would be. For example, mortgage payments might not be considered consumption, and instead consumption might be taken to include, for home owners, an imputed sum for consumption of housing services. Similar issues arise for consumer durables. $\endgroup$ Mar 4, 2021 at 12:56
  • $\begingroup$ @AdamBailey - That's really interesting and I can see exactly how the precise definition would change all sorts of things. If I remember right, that's also why we have all the complexity of VAT - to avoids the definition of "final sale" in a sales tax. If you have any links or ideas for search terms about this being discussed in the 80s, do let me know. $\endgroup$
    – paj28
    Mar 4, 2021 at 15:40
  • 1
    $\begingroup$ @paj28 A consumption tax (albeit termed an expenditure tax) was one among many options considered in the Meade Report on The Structure and Reform of Direct Taxation 1978. See chapters 8-10 and especially Table 8.1 p 151. $\endgroup$ Mar 4, 2021 at 17:54

Consumption taxes could actually be made more progressive by increasing the progressivity of income taxation...

One property of the Marshallian demand functions is that:

$$x^M\big(p,(1-t(y))y\big)=x^M \big( \frac{1}{1-t(y)}p,y \big),$$

with standard notations. It follows that there is a bijection between a (progressive) income tax rate $t(y)$ and a progressive price taxation rate $\tau(y):= t(y)/(1-t(y))$.

There at least two advantages of income taxation: it is quite easy to observe income and difficult to escape income tax. In contrast, price taxation is difficult to organize (there are $J>1$ commodities) and easy to avoid (by organizing your shopping by poorer friends).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.