Skip to Main Navigation
Podcast January 19, 2022

Tell Me How: A Carbon Tax Affects Poor and Rich People Differently

View all episodes on our Tell Me How: The Infrastructure Podcast Series homepage

In this episode, we discuss the distributional impact of a carbon tax and how to design packages that compensate the people who are affected.

Listen to this episode on your favorite platforms: Amazon MusicApple PodcastsGoogle PodcastsPodbean, and Spotify


Roumeen Islam: This is the World Bank's infrastructure podcast. Welcome to our first episode of 2022. Today, we discuss the distributional impact of a carbon tax. Whom does it affect, and why? 

A lot of the discussion around policies to address climate change revolves around a carbon tax. The thinking is that if the cost of carbon were raised to reflect the true cost of emissions to society, then consumption would be curtailed to socially optimal levels.

Governments have to consider not just how to establish a tax, at what rate, and which sectors to tax, but also how to implement it and, in making all these choices, who will end up paying the tax and by how much. They want to know the effect on residents. So, how should they think about a carbon tax distributional effect? Let's find out how!

Good morning and welcome. I'm Roumeen Islam, host of Tell Me How and today I have as my guest Jan Steckel Professor for Climate and Development Economics at the Brandenburg University of Technology and head of the working group on Climate and Development at the Mercator Research Institute on Global Commons and Climate Change. He will be speaking to us today about the distributional impact of a carbon tax. Welcome, Jan.

Jan Steckel: Hi Roumeen, thanks so much for having me. 

Roumeen Islam: It's very nice to have you with us today. So, Jan, my first question is, what was the reason you began your research on the distributional impact of a carbon tax? 

Jan Steckel: We've seen that carbon pricing and also fossil fuel subsidy reforms have been stopped because of massive protests. Sometimes they have turned out rowdy. Understanding who loses from those reforms and how losses can potentially be compensated seems to be extremely important to resolve the political economy of carbon pricing. And even more so, as more and more countries have adopted that carbon tax, others are thinking of doing so, tax adoption and implementation is often a politically fraught area. Energy price changes are socially and politically sensitive, and to basically solve this issue and to provide also academic solutions, we started to look deeper into those questions. 

Roumeen Islam: Yes, it is indeed a very important question because we are seeing how sensitive this issue is both from a social and political perspective and across many countries.

But just to be clear, when we're talking about carbon taxes here, we're not really talking about taxes on all sources of carbon, such as from agriculture or deforestation or methane but only selectively about taxing fossil fuels, right? 

Jan Steckel: That's right. Usually, we only look into emissions from the energy and industry sectors. So, CO2 is emitted when burning fossil fuels. Sometimes in our analysis, at least, we also look into the broader greenhouse gases, but then we look into the agricultural sector, etc., which is very difficult to actually get your hands on when thinking of taxing. 

Roumeen Islam: Is that because the emissions are harder to monitor?

Jan Steckel: That’s it.

Roumeen Islam: Okay, thank you. So, there aren't as many low- and middle-income countries as there are high-income ones that have adopted a carbon tax. And I know that you've done research in some of the developing countries that have adopted one. 

So, what might be material ways in which low-and-middle-income countries differ from high-income countries and ways that would lead to different impacts in those countries? Could you speak a bit about this? 

Jan Steckel: Sure, first of all, we see a large informal sector in most of the low- and middle-income countries, which we do not have in the same regard in most developed countries. 

The second point is the ability actually to raise taxes. So, we know that in many developing countries, the tax rate, as a portion of GDP, is much lower than in developed countries. Then there is, of course, a high ratio of low-income people. And also, the types of fuels that are consumed are very different. So, think of the many people who still rely on traditional biomass, on charcoal, and this is important because this might also interact with actually pricing formal fossil fuels.

Roumeen Islam: These are really very important ways in which countries may differ. We should go into this a bit more later on in the podcast. But before that, could you talk a bit about your research methods and the data you use to understand the impact on households? 

Jan Steckel: Sure. First of all, we need to understand the determinants that are important for our research.

And these are questions like you already talked about, which greenhouse gases do we want to look at? Mostly CO2. Which sectors do we call them? Mostly the energy industry. Which timeframe to recover? Is it the short term? Is it long-term? Is it direct effects also, including indirect effects, for example, that appear through the entire supply chain? 

We, in our research, are mostly interested in the overnight effects of carbon pricing. So basically, think of we go from not pricing carbon to carbon pricing overnight tomorrow. What happens? Who is affected? So, we use mainly…

Roumeen Islam: Sorry, Jan, this is because you think the longer-term effects might be different. Is that why?

Jan Steckel: The long-term effects are much more difficult to accurately determine, and also think of when I'm motivated into this research. It has a lot to do with the political economy. So, what are policymakers really interested in? Is it the generally preliminary effects? Yes, of course, but it is also actually the immediate reactions. Will the reform that I actually apply today, will people actually kick me out of office for this tomorrow because they basically are affected so heavily because they cannot afford to buy bread or food or anything? So, they might get really angry. 

Understanding this is the main goal. And therefore, we use different kinds of methods. For example, we work with household data; we work with micro simulations that are able to look into these short-term effects.

And by the way, I'm personally more skeptical about calculating these more general equilibrium effects. Would you do, for example, do it by computable, general equilibrium models in the context of low-income countries? Because you need to make so many assumptions that, in my eyes, it's just very difficult to justify those, always.

Roumeen Islam: Thank you, Jan. So, I interrupted you. Did you want to go on about what else you have in your models? 

Jan Steckel: Yeah, maybe just quickly, because you talked about the detailed models. So, first of all, basically what we do is we combine very detailed household data, so this is kind of giving us information on how many children households have, what goods they consume, how they cook, what kind of fuels they use, etc. With the so-called multi-regional input-output models. And we need these models to calculate sector-specific carbon footprints. For example, the carbon footprint of the industry sector is more detailed than just the industry sector.

So, the automotive sector, for example, in country X, including all the value chains, also taking into account where the cars or motorcycles in this example are imported from. And if we combine these data and models, we can calculate how much a specific context would affect the households' expenses.

Roumeen Islam: Okay, that's good to know. There are a lot of things that you actually take into account when evaluating the households' decision-making. So, could you speak now about some of the main results and how they differ from what might be found in higher-income countries? So now we're going back a bit to what we discussed a bit earlier.

Jan Steckel: So most importantly, if we do these analyses in developed countries, then a very usual result is that the distributional impact of account price is regressive. That is that poorer households, on average, are, in relative terms, more highly impacted than richer ones. That's different in poorer countries. Here we find that carbon pricing will be generally progressive. So, the richer households are more effective in relative terms than the poor ones. 

But the results obviously depend on the country. All the actual design. That is which sector, for example, is covered by account price. Take China, for example. China has put the emissions trading scheme. So, a price that only covers the power sector that, of course, has different distributional effects compared to also taking the transportation sector into account, or the heating sector, or what have you.

And then, we also find large differences within the income groups. That is even though results are progressive on average, there might be poor households that are highly affected because -for example, in Turkey- some poor household is heated with coal, and they would, of course, be much more affected than households that heat differently, for example with gas or with whatever.

And generally, importantly, in poor countries and across all income groups, we see absolute effects of carbon pricing, that is, without revenue recycling. So just the effect basically, no other measure taken, can really be significant even for the middle class. So, when putting a tax, then the household wealth would at least at the first stage, be reduced.

Roumeen Islam: Yes. And I guess that's important to understand that  So, let's take some of the things you said one by one.

I understand that the distribution of effects are country specific. Now, could you give us some more examples? Could we delve into this a bit more to understand the main reasons which we've spoken about, but could you go into this a bit more? 

Jan Steckel: Sure. So, the

Take vehicle ownership as an example; it increases with income. In poor countries, this leads to richer households using a higher share of income for transportation fuels, and when they get richer, they are more effective, right? Third, the expenditure share decreases again for the very rich because you do not endlessly increase your driving, so to speak.

So, you need a car or a vehicle to get from A to B, but at a certain point, you just get richer, but you don't drive more. So, the expenditure share decreases the gap for the very rich. Also, take biomass consumption, it decreases with income. So instead, people use more of their income for cooking, and -depending on the country- also for heating. However, biomass is not taxed. Forms of fuel such as kerosene, LPG, or coal etc. are. So richer people are more affected.

Roumeen Islam: Let's pause a bit here. You're saying that poor people will suffer a welfare loss, but even if they do, one would like them to consume less. Let's talk about this a bit more.

Jan Steckel: There are various perspectives to that. First, from an equity perspective, we don't want very poor people that already have a very low income to suffer, right? So, this is even more true as We know that it has negative effects on gender, on nutrition intake, as refined in a recent, yet unpublished, study on Uganda. But these aspects need to be considered when thinking of carbon pricing in poor countries. 

Second, there's an acceptance perspective. So, , such as for example, Ecuador as I read recently. And these protests have the power to stop these reforms.

It is important to know who those hardship cases really are to communicate to them and to think of how to compensate them.

The differences within income groups, or what we call horizontal effects, are more severe than the differences in between groups, or what we call vertical effects.

Roumeen Islam: Now, could you just speak generally about what's one way to design a successful reform.

Jan Steckel: I think for designing a successful reform, it's important to go beyond the standard economist toolbox. So, we often prefer to make a lump-sum transfer of the climate dividend, the same amount to everybody. However, countries might have institutional difficulties setting this up: how to transfer money, how to reach the people in need. Therefore, we increasingly also think of using the existing targeting mechanisms. 

Roumeen Islam: All right, What are your views on using tax reductions to compensate for a carbon tax in developing countries?

Jan Steckel: It's true. The standard literature highlights that revenues can be used to lower other distortionary taxes, such as taxes on labor. And that's sincerely also a very good idea, but in developing countries, I'm less sure, honestly. So generally, those countries have a much lower tax income for GDP, they have difficulties raising taxes firsthand. So, it's probably better to use the revenues to invest in infrastructure. Think of electricity, think of improved access to stoves, for example, or water, etc., or transfers. So, I think for developing countries, those two options are probably preferable than actually cutting down on the existing revenues that are already little.

Roumeen Islam: That's an important thing to think of. You're right. Sorry, did I cut you off? 

Jan Steckel: No, I just wanted to highlight that I forgot to say previously that

There's little research in developing countries and that, but

Roumeen Islam: That's an important point that you just made for earmarking, which, one might not necessarily think that taxes should be earmarked.

Now, does the impact depend on how the carbon tax is designed? 

Jan Steckel. And so, I think, for example, on the sectoral coverage. So, many countries, when they start discussing to put a price on carbon, might not firstly think of the entire economy.

They might start with the power sector or the transportation sector or the industry sector. And that is fine, but like all of these central designs also imply differences in the distributional impacts. And I think it's important to have this in mind. 

Generally, what  So, have in mind that , and I should think how it is perceived fairly. This is in particular true if we think of poor societies, who, of course, have very limited emissions, who are not responsible, etc.

Roumeen Islam: Yeah, those are all excellent points. And I, in particular, also like the point you made about the perceived fairness of reforms: the actual, as well as the perceived fairness, both matter in this regard. 

So, Jan, could you speak a bit more about the effects of the carbon tax, depending on whether or not how informal, how large the informal economy is?

Jan Steckel: Yeah. Usually, we have a problem in developing countries that these informal activities cannot be taxed. For the carbon tax, that is different because even if you are involved in informal activities, you usually still need to buy fuel, for example, on the formal market. This fuel is relatively easy to monitor, like how it is used, how much it's used, and therefore it can also be taxed. Therefore, putting a price on carbon is also a possibility to indirectly at least tax those informal activities. 

Roumeen Islam: Okay. Thank you for clarifying. Before we end today, is there something you'd like to add, Jan? 

Jan Steckel: I would like to add on one point because, for some people, it might sound strange to discuss carbon pricing in very poor countries.

Why do I still think that is twofold.

First, I do think that The second is that I do think that

So even if emissions are arguably very low, also historic emissions are arguably very low, I think it's just important to set the price signals in the correct way to ensure that development is going in the right direction. 

Roumeen Islam: Thank you very much for an excellent discussion in which we all learned a lot today. Thank you.

Jan Steckel: Thank you so much. 

Roumeen Islam: Well, listeners, what have we learned today?

Firstly, For example, the size of the informal sector, the types of fuel consumed by rich and poor people. The effectiveness of the taxation system. These are all factors to consider in assessing its impact in a country.

Secondly, And this is done mainly through direct transfers, reductions and other taxes or improved services through safe infrastructure investment. However,  nor to implement the appropriate fiscal policy.

Thirdly, If you'd like to learn more about the carbon tax, do listen to our episode eight, tackling the social cost of carbon. Thank you and bye for now. 


Blog: What a carbon tax can do and why it cannot do it all

Website: The World Bank and Pricing Carbon

Report: State and Trends of Carbon Pricing 2020

Data: Carbon Pricing Dashboard