 Welcome to this session. It's entitled Reforming Fossil Fuel Production Subsidies. I'm Laura Merrill from Global Subsidies Initiative of the International Institute for Sustainable Development. Everybody will get into depth with fossil fuel production subsidies in their presentations, but I'll just say that subsidies to production matter, and they matter in their existence and in their removal. It's basically because they're encouraging more production of fossil fuels by making them cheaper, just making them cheaper to extract and to export. Really, combined with consumer subsidies, they make it very difficult for us to disentangle ourselves from fossil fuels and from a world where our development's based on carbon. So, how do we extract ourselves from these fossil fuel subsidies and upstream at the producer end? And how do we remove these subsidies to production? Some of the questions that the researchers here have asked are, what are fossil fuel production subsidies upstream to oil, gas, coal and electricity? How are they different from and interact with consumer subsidies? What impacts do they have on the environment on society and on carbon and climate? And how can they be removed? What will they be replaced with, if anything? So, we're really privileged to have panellists from Russia, from Kenya, from Peru and the US, and I'm going to introduce everyone in one go, and then everyone gets 12 minutes to present. It's the first speaker is Yvette Gerysim. She's senior researcher for the Global Subsidies Initiative at IISD. She's a friend and colleague of mine, and she's going to be presenting on unlocking supply and locking in carbon, the paradox of determining which fossil fuel subsidies are the worst for the climate, but I think we'll just start with you and then we'll come back to the others. Introduction, Lauren. Good afternoon everyone. Fossil fuel subsidies have been long described as a low-hanging fruit of climate change mitigation, a low-hanging but prickly, so that's why I have an image of a prickly pear here. Well, a low-hanging because obviously you don't have any costs when you implement this policy on the opposite you get gains, you save public money, and prickly because of the political economy challenges. So, our spare is distinguished audience, the definition of fossil fuel subsidy because there are people in this room who have worked on the issue for more than 20 years. I have worked on this for six years, and I can say from this experience that there is one thing in common between a fossil fuel subsidy and pornography, which is sometimes very difficult to define also legally, but when you see it, you know it. So, I'm doing this presentation together as part of research with Andrea Bassi. So, we're working on, and together with some people who are also in this room, we're working together on a report which is looking at the climate benefits of fossil fuel subsidy reform on the produce end and it's looking at such benefits on a global level. So, what I'm presenting is work in progress, it's the basis of assumptions we are doing also for quantitative modelling. And the two key questions we had, one is what are the worst subsidies for the climate and if these are the same subsidies that should be phased out or immediately removed first, and these are two questions which are quite practical for campaigners, for looking to award a price to the worst subsidy for instance, or to policy makers who want to deliver on the commitments under the Paris Agreement. So, how much are producer subsidies in the world is something we know quite a bit about, but not everything. So, GSI, a few years ago produced a heroic estimate of 100 billion globally. Then the work that has been co-published by OCI and ODI and where GSI also contributed case studies estimated just within G20 producer subsidies at 70 billion and here's a graph. So, for our modelling exercise we try to aggregate these numbers by fuel. It's by itself not a trivial exercise because a lot of subsidies are cross cutting, a lot of them are going to both oil and gas and it's the same industry effectively or some of them applied to all fossil fuels and on top of it also to all minerals. So, it's difficult job in itself just to measure, to quantify and disaggregate fossil fuel subsidies. We have to make some assumptions which we have done for instance, we attributed numbers based on the role of each energy type in the primary energy production in the country and it shows you more or less that the more capital intensive the industry is like oil is the most capital intensive industry, the more subsidies it gets and on the opposite coal is less capital intensive so it's showing up as a lower number in terms of subsidies. The climate impacts of fossil fuel subsidy basically depend on three groups of factors. One is subsidy design so because subsidies are there for a reason to stimulate some kind of activity. There are also unintended impacts of these designs but this is something the governments can control. The other group of factors is something the governments can't control is the physical characteristics of projects, is the economics of the projects and then you can say in rough terms although I'll do some caveats later about this. So, there are by their nature projects that are viable only with subsidies so that's where we are looking at the effects of unlocking them into production. There are projects which are viable even without subsidies so subsidies arguably have no effect on them and finally there are projects which are unviable even with subsidies. And the third group of factors here is everything else and that's what makes subsidy analysis and analysis of climate consequences of these policies quite difficult because there is such factor as well price for instance which completely changes the economics of projects. There are technologies which we discussed and they can also influence the economics and competitiveness of different projects in very drastic ways. So there is a lot of uncertainty in general. So if we talk by design because this is something the governments can control, we can talk about subsidies that give a signal, a very long term, a very strong signal to investors, to companies. And this is why for instance maybe some of you have heard before G20 group of insurers came up with a call to the governments to stop fossil fuel subsidies because subsidy gives a very strong signal to the industry to invest in particular type of assets. And this of course has implications for risks and insurance costs. So then there are short term signals, more of temporary nature so it's in terms of cost reduction. So the first group of with very long like beep beep signal is the type of subsidies that reduce fixed costs at the very beginning of the project. So the short signal comes if it's the reduction of operation costs and also consumer subsidies fall in this category because of course from the sales perspective you can only start selling the product once the project becomes operational and it's longer into the future. And finally arguably there can be subsidies that give no signal to the industry and I'll come back to this later. So I already mentioned this time value of subsidies, time value of money. I'll not go deep into detail on this but basically capital in the industry is scarce, investors are very myopic, they are looking at the present value of projects. And this value depends on discounting and then you have the tyranny of discount depending on which discount rate you select. You will get a different value of a dollar, in this case a dollar of subsidy in this present year or 10 years from now or 20 years from now. So very roughly if you take project life cycle in the industry at 25, 26 years, you take five years for project development, this more immediate costs and then you have all the costs coming from the fifth year to the 25th, 26th years. So you have a difference by a factor of three if it's the discount rate of 10% and it's a factor of two and a half if a discount rate of 8%. So this is the cash flow of the project, very typical coming from a textbook of the oil and gas industry and discounted in this case. So we have the majority of costs here in the negative coming at the beginning of the cycle and that's where capital costs are incurred. So that's why basically the strongest signal which is sent by governments is if companies for instance get accelerated appreciation, if they get free infrastructure in terms of pipelines, in terms of roads. So this type of subsidy will unlock the investment into projects immediately. So the moderate signal, the big signal comes at the latest stages when the project starts producing. So it either reduces the operational costs of running the project, for instance labour costs or health insurance costs for coal miners, things like this. Or it will increase the revenue at the producing stage for instance by guaranteeing the market for oil, gas or coal-fired electricity. And that's why consumption subsidies can also have a local effect on high carbon assets. And finally, theoretically we can say there are projects, there are subsidies which, there are projects which are viable without subsidies and we at JSA have done a comparison of two projects for instance in the Arctic part of Russia. So with subsidies one project became commercially viable and went ahead. The other project we looked at, the pre-aslamnir platform was actually viable even without subsidies. So you can say subsidies have no impact on unlocking new capital, unlocking new assets. In this case it still increases the revenues of the industry. You can ask questions what about the opportunity cost, what if other industries receive the same level of support. You can ask questions about the investment of these profits, recycling of these subsidies within companies. So it's quite precarious to say that some subsidies have no impact. Although in theoretical framework like we are doing for macro-moderning, of course we have to differentiate. So there is also frequently asked question about coal subsidies, are they worse than oil and gas subsidies? Short answer is on production side no because everything depends on the characteristic of an individual project. The percentage of costs of said by subsidies and as we said in the coal sector the capital costs are lower than in oil and gas. And then a lot of coal projects are viable even without subsidies. So which brings us back to, okay so we have some kind of ranking of subsidies we can say some of them are possibly worse for the climate than others. So can we just eliminate those and that will be already a step forward. So can we just cut one quickly pair and feel good about it. So unfortunately there is leakage effect and what happens is that once you cut one the political economy can make subsidies reappear in different places in the same country or in a different country. So in this case a call for elimination of certain types of subsidies is much weaker than a call for elimination of all types of subsidies. And that brings me to conclusions that if you want to answer this question about which subsidies are the worse. Of course the best is to look at project level but not everyone has this luxury and then you have to do a trade off between detailed analysis and assumptions. That prioritisation is possibly still not the best way forward because of political economy challenges and also because of economics such as leakage, such as fuel substitution. And the best call for those who want to eliminate fossil fuel subsidies is the call to eliminate all of them to all types of fuel and both consumption and production. We are releasing the new report around Marrakesh. So if you have any questions we are very much looking to your feedback. Thank you. Thanks Avetta and now we're going to hear from Kennedy Inverva. He's a research fellow at the African Centre for Technology Studies Act based in Nairobi in Kenya. Welcome Kennedy. Thank you Laura. Ladies and gentlemen, I'll be making a presentation trying to explain the variation in fossil fuel subsidiary forms across country because a case has been made for fossil fuel subsidiary forms as an important climate policy tool. But we know if we apply it across both do we find some variations in some countries or not? So to give a brief outline of my presentation so first have a brief overview of fossil fuel subsidiary forms and how they've been related to climate policy. Secondly I'll share the research questions and the methods that I use in my analysis. Finally share the findings and some conclusions from the research. So the main focus of this research was on looking at fossil fuel subsidiary form as a climate policy tool because there have been many arguments that have been made about using it for instance removing fossil fuel subsidies. The G20 for instance is looking at inefficient fossil fuel subsidiary forms so the debate is emerging towards adopting it as a climate policy instrument. But I think as you've heard from much of the research a lot of research has focused on the demand side and not much on the supply side. So this research will look at supply side of course and also look at the conditions which might lead to those reforms taking place. So there's a lot of theorising so trying to generate a broad theory to try to explain so as all theories maybe in some cases it might work in others it might not. So I stated the main question was looking at under what conditions can fossil fuel supply side reforms be successful or unsuccessful. So the main focus was looking at time to an event and time to that event is undertaking a reform. So I developed an original data set and the main variable that I was looking at dependent variable was how long does it take for a country to undertake a reform and by reform I mean developing a national legal framework whether it's a law or policy that has direct impact on energy supply and climate policy. And I got a very good data from a database developed by the London School of Economics called the Globe Data Set and it has national legislation laws and policies that actually have a direct relevance to climate change whether it's adaptation or mitigation and most of the national laws and policies that I looked at all the I think I looked at 95 countries which produce fossil fuels many of them have to deal with energy supply. So that was a very convenient data set to construct trying to look at the different factors that will lead to that event happening which can be a very good proxy to a reform happening related to energy supply and climate policy. And so the main hypothesis that I had is that the higher countries preference to climate policy they had the probability of reforms. And by measuring climate policy preference I looked at the number of laws in that database that a country has. So if a country has a higher number of laws then there's a higher probability that it will undertake a reform through a law that is related to climate change. So that was a proxy. And I also looked at other factors that have been put forward in literature. And I looked at political for instance like do the different types of governments or governance have an impact will reforms be maybe more successful in countries that are democratic or less democratic or countries that have a federal system or less of a federal system or a centralized system. And I also looked at the economies that the size of an economy have an influence on maybe if a country will have a higher probability of taking a reform or not. So those are the three main hypotheses I was testing. And of course the statistical analysis I used was a time to event model trying to measure the time to that event which is a reform happening. And I took the years 1992 to 2013, 1992 because it was a watershed year for global environmental politics. And many of these issues gained a lot of traction especially on the international front. And 2013 because most of the data that I got was limited to 2013. It would have been better to go maybe all the way to 2015 but that was a bit challenging. So one of the findings I took from that global database, the national laws and policies for those 95 fossil fuel producing companies are countries looking at what is the cumulative number of those laws for those countries over the years. As you can see there's been quite a steep increase so some of them might be positive or others negative but that can show you some activity in the climate preference. And these laws include adaptation and mitigation but it is a good proxy to look at the preference overall. So these are the main tables that had the key findings. So one of the key things was one we looked at a growth of greenhouse gas emissions in percentage like looking from 1992 because it's been arguing in literature that a climate policy, if you have a fossil fuel subsidy reform as a climate policy tool it's a cheap or a low cost tool. So we tested that hypothesis and found that there was evidence although weak that it can be used as an option, one of the low cost tools. I'm not an economist but that was on the main arguments. Secondly the number of climate related laws, national laws, these measures the preference to climate policy. So there's a very strong evidence that countries that have more laws and policies that are related to climate are more likely to enact reforms. So when I look at carbon emissions to countries that have higher emissions maybe consider using reforms as a tool to address those. So the theory suggests that that will happen although with weak evidence. So one asterisk is for 90% confidence, two asterisk is for 95% confidence while three asterisk is for 99% confidence. So looking at the politics aspect of it I didn't find sufficient evidence to that's a post that the variation in the types of government or governance actually has a strong impact on leading to reforms because if you look at for instance like democratic governments by whatever definition you use you'll find that there's a lot of representation of many actors and that's where a lot of contestation happens. You'll find people who are very forward looking pushing for climate policy but you'll also find that fossil fuel companies represented there. So there's a lot of contestation. When you also look at autocratic governments you'd expect most of them based on literature have been very antagonistic to climate policy mostly because they rely on fossil fuels. So the probability or chances of them undertaking reforms is almost there. So I didn't find much evidence to support it in this broad theory. So looking at the economy aspect of it I found evidence that an increase in the size of economy would increase the probability of the reforms happening because when you look at the GDP maybe one of the explanations might be that big economies are more diversified and maybe they have access to alternative technologies to replace that fossil fuel technology. So that might be one explanation and it was actually very strong for developing countries. When you look at the figures in the parentheses they represent developing and developed, least developed countries because I only heard from the 95 countries 11 least developing countries which is a pretty small number to some degree. So I combined it with the number of developing countries which are 52 in the sample and you can see there's a very strong evidence that an increase in the size of economy would maybe reflect the diversity or availability or technologies to some degree. And just a note a number higher than one shows a higher probability like an increase in that factor leading to higher probability. A number less than one means the opposite. So another interesting thing was looking at official development assistance. So I specifically looked at that variable because in the past development institutions have been channels for passing through norms, different norms of development sustainability etc. So I looked at the percentage of countries income that comes from ODA and it was very interesting to find that an increase in ODA leads to a higher probability. Maybe it's because of the channeling through norms and the influence and leverage that those development institutions or rather those aid institutions have especially on developing and least developed countries. And finally I looked at the fossil fuel rents as a percentage of the income a country gets. And you can see there's a significant evidence especially for developing countries that the more they rely on fossil fuels for their income the less likely they are to undertake reforms. So moving on to the part that answers to the first question that I posed. You can see that in blue is a combination of the developing and least developed countries and in green is a developed countries. And you can see developed countries are exiting from the sample really fast because that shows like they're making a reform and leaving. So developed countries actually have a higher probability of undertaking these reforms than developing and least developed countries. And you can see this for example through the OECD, through the 20. They are the ones who are sort of like the fast movers in these issues so we'd sort of expect that. So finally back to the question. So the question we looked at under what conditions can supply side fossil fuel subsidy reforms be successful or unsuccessful. And my broad theory suggests that countries with a higher preference for climate policy are more likely to undertake reforms. Secondly the type of government does not have a major significant impact on other reforms will happen or not. And thirdly countries with bigger economies are more likely to undertake reforms. So putting this in the broader context of the discussions today brings about the question like one. There are countries which depend on these fossil fuels but if you tell them to live in the ground what is the alternative. Secondly the same countries have very ambitious climate plans especially in the INDCs. How do they reconcile the two? So it opens up again the issues that were debated before. How do you have a just transition? How do you have an equitable transition because this entire like the processes varies across different country groups. So that is the main presentation of the theory. Of course there are many other interesting variables that could be considered. But just to test the broad ideas, these were the ones that were considered in this time. Thank you Kennedy. Next up is Sue Clarie, a radio guteris. She is from the Environment Ministry from Peru with Peru. And she's going to be talking about measuring the impacts of eliminating subsidies and assigning taxes to energy products in Mexico through a general equilibrium model. Better. Thanks. I'm going to present one of the results of my PhD dissertation that was about the Mexico situation about avoiding subsidies, specific production subsidies and also carbon tax and transport tax. For Mexico it's important to be one of the countries that develop different climate policies and activities about mitigation and adaptation of climate change. Also we have a climate change law with an aspirational goal to reduce emissions 30% for 2020. So we have the aspirational goals but how we are going to develop these activities or how we are going to reduce emissions. So they propone different tools, MRV system, national emission register, that is a specific difference in the case that they are just going to know about the emissions and they are not going to do transactions or a cap and trade in the register. So the other part is we already have energetic reform. We are trying to change to having just one oil company, two different companies, private sector, investing in oil and gas, then a fiscal reform when we include a carbon tax. So this is the national context. Also this year we ratified the Paris Agreement so we have more commitments to follow in the next year. So the idea is to know what about production subsidies. These are not important at all because most of all think about the consumption subsidies. They are bigger, these are more important for us and the production subsidies are also important in this case. I developed a general equilibrium model considering the full economy. We also consider 10 different types of consumers and we are not considering the benefits of emission reductions that are the result of avoiding these subsidies. Another thing that is important of this model is that allow us to compare between different sectors and also between different policies. Most of the time we have specific models for each sector or for each policy and in this case the model allow us to compare the different sectors and policies with the same database. So I consider the sectors that are related to oil, gas and electricity. The producer sectors and also the sectors that put this different input for their activities. Well I developed two different models. One is a static model. I'm just going to present some of the results. And why it's important for us to consider the production subsidies. If we can see these are the amount of the production subsidies and these are the consumption subsidies. So it's a small quantity compared between both but also these amount are the same that some social programs in Mexico. And it's important to consider that we need to invest more in education and health or even in poverty programs. So these are not just little quantities. They are important for us if we want to invest in other programs that help the country. The results about eliminating these subsidies. We have the results in three different variables. These are emission reductions. These are the additional income to the government. And here is the welfare of the consumer considering a equivalent variation. To say electricity has a significant impact if we avoid the subsidy. It's less than the 1% but it's more important in the emission reductions also in the government income. And if we consider the part of the consumer welfare we have just when we avoid all subsidies the 50.50%. So it's not a big amount the impact that we have and also we can use that money to another programs or other activities. The other path was to a second in taxes into different points. One is trying to simulate a carbon tax that it's different for each kind of food considering the content of carbon. And transport tax that it's just for the products that we transfer from another cities to the capitals to provide food or different vegetables. So the government income increase in the carbon tax more than the road transport tax. And also when we consider the consumer welfare it's bigger the impact with a carbon tax than the transport. The important thing here is to consider that the lower the size have more impact with the carbon tax than the difference between the transport, the first size to the last one. So the carbon tax sometimes will consider that it's better because you punish the people who use their car. And in the Mexico case not exactly that. The people who don't have a car use public transport and everything and it's expensive for them if we assign a carbon tax than another kind of taxes. So this is the part of the static model and also we develop a dynamic model considering just the different scenarios about taxes in a period of changes. This is the result for the whole period in the principal variables that are emission reductions, government income and welfare. So first if we assign taxes from 2% to 10% we don't have a big emission reductions or bigger difference between them. But if we want to have important emission reductions like 3% is when we assign a 30% tax, carbon tax. So that's the first one. But also the impact to the welfare is 3.5. So we have to consider what is better or worse. Also the government income increased just around 2% and this gave us the opportunity to make transferences to the first 5% of the population. So maybe it's an option to consider high taxes and also to make transference to a population. It's one option. And well the general results of all the different things with it shows the same. The more important thing is the government income increase more when we have a bigger tax considering the consumable worker. It's a bigger impact for them if we remove the gas subsidy and not the other subsidies. And the emissions will also add tax on oil production has the better impact in the reduction emissions. Well some reflections for us is that if we want to reduce emissions have to consider many factories, not just fiscal policy or an energetic reform that we need. Mexico is an oil dependent country so we have to change a little bit our energy metrics. So also we can invest the amount that we have with taxes or the amount we have without eliminating subsidies to a new technologies. Also renewable energies or the other thing that is important is that the carbon tax has to be focused to something to do. It's not just another new tax that we can use for all the things that we use like health or everything. We have to focus that amount in specific activities. So that's the only way we are going to go to our aspirational goals of 30%. So that's all folks. So please welcome Doug Coplow, founder of EarthTrack, who's come and will be talking about fossil fuel subsidy reform in the United States, impediments and opportunities. Thanks very much. It's nice to be here. I want to cover three things. One, a little bit of discussion of the history of fossil fuel subsidy reform in the US and go a little bit into the impediments that we faced and some ideas for addressing them. And then also to introduce some work that I've been working on with SEI that's trying to unroll the subsidies at the field level and see how it changes the investment returns. And we're fortunate that we have Mike Lazarus, Pete Erickson and Adrian Down in the audience. So if there are questions that come up I can't answer they're here to help. So we've been dealing with this question for a long time. This political cartoon is actually from 1889. The fat guys in the background are actually money bags for big industries at that time. There was a big concern that they were intervening with the Congress, getting control of government to direct money to the individual sectors and keeping the public out, which is pretty much what happens with oil subsidies. A lot of these industries have gone by the wayside, but the ones with yellow arrows, we've got coal, oil and gas, they're still around. So that's 1889. This here is 1927. OK, so one of the big first tax breaks for oil and gas in the United States came in in 1924 or 1925. It was a percentage depletion allowance that allowed firms to deduct from their taxes more than they actually invested in their oil wells. Only a couple of years later we had the division of investigation of the joint committee on taxation kicking in with the report, trying to assess why was this happening, was it fair, was there any basis for these subsidies. In fact there's a great quote down here from the senator who sponsored the legislation acknowledging there wasn't such an economic basis for it. In fact they plucked 27.5% out of the air because it was big and they wanted a lot of benefits, but also because it seemed scientific because it was so precise. So this type of manipulation of facts I think is something that anybody who tries to deal with subsidy reform addresses on a regular basis. This isn't to say that we haven't made improvements in the United States. We actually have on tax and tax expenditures 1974. That's quite a while ago. We made federal reporting of tax expenditures mandatory. This was actually picked up over time by most states in the country. The problem was the states don't always assess it in exactly the same way, but they were assessing it and that's a good thing. In fact the government accounting standards board has now issued a formal rule making on tax abatement disclosures which will kick in next year. So we will have standardized reporting required for all levels of government in the United States. That's a big improvement. Credit subsidy reporting. Again we started early 1991 federal efforts to have credit reporting in the budget kicked in. It was complicated. The states really haven't stepped up in this area and even the federal metrics are not very good. So they ignore the cost of administering the credit programs, but most importantly they benchmark it against the treasury's cost of borrowing which is very low risk capital. And what they should be doing is benchmarking against the riskiness of the ventures that the government is supporting. So a lot of room for improvement there. We've actually got some of them in the United States, not as many as other countries in the energy sector, but we do have things like the power marketing administrations. When there are discreet commercial activities by government owned entities, we've got pretty good rules. So we require standard financial reporting. It has to be independently audited. Those audits are made public. There is accountability. But there's a lot of blurry lines of government involvement. So we have the federal government maintaining the inland waterways and it just so happens that 50% or more of the tonnage that moves through the inland waterways is oil and coal. We have a strategic petroleum reserve, which is billions of dollars of stock oil. These are blurrier issues of state owned enterprises and the data is pretty bad. They don't do the accounting properly and therefore we often don't know how much we're spending to provide these goods and services to the fossil fuel sector. There are some positive trends in the US for sure. Ten years ago it was kind of the line agencies, Environmental Protection Agency, Department of Energy that were focused on energy subsidies kind of from the policy angle. Today it's very much coming from the treasury and the president. This is a pattern that I've also seen happening in other countries. The Obama administration put in fossil fuel subsidy reforms in most of its budgets. None of them got through, but they were putting them in. In growing data, OECD, other international agencies, NGOs, some of them are here, oil change, overseas development, doing a lot of great work to have more frequent reporting, broader coverage of countries, broader coverage of the US, and that's been a great help. Even the G20 is moving a little bit, a little bit more than I thought it was going to. We'll have to see how that plays out. Ten years ago or 15 years ago, when you talked about the importance of removing fossil fuel subsidies, there was sometimes a question as to whether people thought it was important. But now it's very broadly viewed as a critical plank of any type of greenhouse gas reduction strategy and I think that's a very positive improvement. Yet when I polled my colleagues on the issue of how we're doing with subsidy reform in the US, I got universally negative responses. I mean things like you got me, what reforms, or in crying lines mud and no, I can't remember the last time the US successfully reformed a fossil fuel subsidy. So this is not optimistic stuff and it's kind of in line with my view of the situation as well. In some narrowing over time of subsidies to oil and gas, some reduction in the levels they can get, some restrictions on international operations, or the largest companies can't tap into everything. But even in the very broad based tax reform of Act of 1986, most of the oil and gas subsidies survived when lots of other things were cut. And the industry is very active and is always looking for more. So we see refinery expensing, the use of the IRS private letter ruling process to expand their ability to use master limited partnerships to escape corporate taxation. We've seen royalty gifts on the Gulf of Mexico, big push now to have large subsidies for carbon capture and sequestration in the US. Even though most of the captured CO2 is going to be re-injected into oil and gas wells to pull out more oil and gas. So this is a very challenging process. We've seen more success on the administrative fronts with leasing and royalty modifications, regulatory improvements, particularly following accidents. And we've seen a couple of cases where people have figured out how to use tax subsidies in interesting ways and the cost ballooned to billions of dollars. And Congress was actually embarrassed and got rid of it. But these are far and few between. Part of the problem I think is numerical. So in the lower left corner is a little smiley face. That's right at the zero. API is the American Petroleum Institute, which is the largest trade association in the US. And according to them, there are no subsidies to oil and gas. The official estimates are EIN Treasury, a little bit higher. Then we've got OECD's work and OCI higher still. But there's a pretty big spread here. So clearly people are using different definitions in what they're picking up. And we can see that as well that when you break out these numbers by type of intervention that the spending and tax expenditure bars over the left here, so grants and tax breaks, is a vast majority about what we're picking up. If that were the only thing that the fossil sector was getting, fine. But there's all these other ways that governments transfer value to private industry that we're simply not measuring very well. And I think one of the challenges that we face going forward is how do we start picking up these things so we can see the full picture of what's going on? This is just another view of that same issue here with the concentration of direct spending and tax expenditures. Another important lesson I think is the importance of subnational subsidies. OECD has done tremendous work. Jehan back there knows the suffering that it took to get it. To go through all the states and provinces to try and tease out what's going on at these lower levels of government. It turns out it's pretty important. In the U.S. about half of the measured benefits are at the subnational level. And we need to do a better job making sure we build that out. Why are we having these struggles when clearly the power of the groups we're up against is a big issue here? Our friends at the American Petroleum Institute have a budget of almost $250 million a year that dwarfs even some of the largest environmental groups in the United States. So oil change is doing tremendous work improving transparency here but easily outspent by the big guys. And they're not, the trade associations like API really aren't interested in the earning. It's a political organization and this is their head tax policy guy basically coming out and saying all these things that we get everybody gets them. They're not really subsidies and it takes a lot of work to try and contest that process on an ongoing basis. The other important point is that we sometimes focus on subsidies that are specifically targeted oil and gas. But the guideline here needs to be subsidies flow to power. And if you're in a part of a country where the fossil fuel industry happens to be powerful you've got to look at everything. This is a review I did of Hurricane Katrina tax exempt bonds. Okay, Hurricane Katrina hit the southeast United States. It wiped out large parts of a couple of states. Congress put through a whole bunch of money to help them rebuild. Nearly two thirds of those bonds were captured by the fossil fuel industry. And that's a great indication of the challenges that we face and the importance of not narrowing what you look at too much. Now I want to shift to some of the work that I've been doing with SEI. I've got these all marked draft. Why are we coming up with draft stuff this meeting? Number one, because it's cool. We want to talk about it. Number two, because we actually welcome input from you guys. So if you see things, if you see ways you'd like us to present the data differently we would welcome that input. So what we did here is we took data at the project level on the projected production levels and economics of oil and gas fields around the country concentrating in a number of basins. And then we synthesized removing the subsidies to see how it affected the rate of return. And the baseline rate of return on average was about 7.7%. Over here is the internet credit which is widely viewed as the most significant subsidy to oil and gas. But what we discovered is there's some little stuff that no one pays attention to that actually matters quite a bit. The yellow bar in the middle is actually the excess road damage that heavy fracking trucks due to the roads in Texas. And it turns out Texas cares about this. They've done some pretty detailed studies. It's about $4 billion per year. So when we run it through our models it increases the baseline IRR by about 1.9 percentage points which is a boost of about 25%. So this is a big deal. And again it illustrates the need to go to the subnational level and to look carefully before concluding that there's no problem. This is a great map that shows the field by field level data where the blue dotted lines on the bottom are what the return is on these fields with no subsidies. The fat orange dots, I'm sorry it's a little hard to see, are the projects where they went above their hurl rate because of the subsidies. So I've kind of tagged these things, the leakage zone which is over to the left projects where public money is going in and it didn't really make a difference in the outcome. So it's sort of largely leaking away to the people who own the fields or own the resources. And what I call the abetment zone. See carbon abatement, carbon abetment. It's not actually trademark but I just want to say if it catches on you heard it here first. So this is an example where taxpayer money is unlocking greenhouse gas emissions, potentially significant greenhouse gas emissions that otherwise would have stayed in the ground without the subsidies. So you can see that it's sensitive to the price, the more expensive the oil is, the less important subsidies are. But along the boundaries of today's price is about $50 per barrel of oil, roughly half of the new oil that's not producing is dependent on subsidies. So this is a big deal. There you are. So this chart is a little bit less beautiful but it's basically saying the same thing. So 47% across the US is subsidy dependent. We tried to ballpark what that might mean in terms of carbon emissions, about 9 gigatons of CO2 within the US. Less obviously if you look at the net reduction because there's some foreign oil that will flow in as you start removing these subsidies. But still it's a significant lever point to look at when we're trying to deal with very difficult challenges on climate change. So finally where do we need to go? We got to fill in the data. We've got to have the data more quickly so that we can be responsive to the policy debates. Be more relevant in decisions as they're being made and ensure that we have fiscal alignment with our environmental goals. And also this idea of policy contestability. So we always hear justifications for existing subsidies that say it's supporting development, it's supporting jobs. A lot of times there's lots of ways to bring jobs or to bring development to regions. We need to be able to come to bear more quickly and say you're choosing a pathway of subsidies to energy that's causing all these problems. But actually there's all these other things you can do instead. Your choice for your state, for your region or for your country. Thank you very much. All right. Thank you Doug. I mean those are all really interesting presentations dealing with policy with Mexico. I saw in their NDC, their nationally determined contribution. They've included energy sector reforms actually and there has been reform recently. And that huge figure from Doug there linked to oil and subsidy dependent oil, which is extraordinary. And the fact that your research didn't find so much Kennedy, all these very strong links. But one of the points that came out from one of the other presentations around government revenue savings and sort of social impacts. And then the final third thing sort of being this code benefit of emissions. So perhaps it's not just about climate policy and it's a bigger picture around government policy and so on and so forth. But we should open this up to the floor and take questions. And let's take four to start with. Okay. Up here. Thank you Henri Weisman from Idry. Just to follow up on what you just said because this is a question to Kennedy but that echoes a lot for example with your very last work. Your starting point, your assumption that there is a direct connection between climate change and reform of subsidies. Which I'm not completely convinced of. I'm not convinced that climate change is the primary reason for reforms of subsidies. Which I mean what we see there are some examples like what happens in Indonesia for example. Where the main reason for reforming subsidies is that it's regressive. Just it's not an efficient way to reduce the inequalities just because the richest consume more and so benefit more from the subsidies. And when under a constrained budget it's a very inefficient way for the governments to use these resources to reduce what is their main concern which is inequalities. And then the emission reductions is obviously a benefit of that. But I'm not sure that it's a primary reason for the governments to take that into consideration. To go to your analysis I wonder whether you could reframe it to start differently with not climate change as an entry point. But some other dimensions at least that you could include in your regressions in order to take that into account. I'm Fernando Tudela from Mexico, the Center for Global Change. Building upon the last presentation I think we have a fundamental difficulty there because we are not just measuring in different ways but we are measuring different things when speaking about subsidies. And there seems to be a lack of symmetry in the positive taxation taxes. We all know a tax when we see one. It's legally framed in a way but there is a lack of definition in terms of what the subsidy is. We saw data in G7, G20, APEC, OECD, IMF, you name it. They have all said that we have to phase out the subsidies. And nothing much has happened and this is an understatement. But even if you consider IA, National Energy Agency and OECD, which are sister organisations, they have different ways of measuring subsidies. IA is focusing on the difference between the domestic price and the international price. A very simple way of defining subsidies although extremely restrictive in the extent. And while OECD is achieving some kind of bottom-up consideration of different programmes that may help production of fossil fuels. You go to the panel. The question of the panel is, are we ready for a discussion on achieving a shared definition of subsidies so that we can measure the subsidies in a harmonised way across countries? Is it feasible? Would it make sense in the sense that we cannot manage what we cannot measure? And do you think collectively, the panel, we can achieve that, a common definition of what the subsidy is? Is there any other questions in the audience? OK. It's more common to respond to what we just said. Just to say that I think people tend to underestimate the extent to which there's agreement on definitions. First, most countries are signatories to the WTO agreement on subsidies and countervally measures where article one has a definition. So most countries have signed off on the definition of a subsidy. And secondly, in the context of the G20 peer reviews, the countries that participate have signed on terms of references that define a subsidy and it's pretty close to the WTO definition. And last, just to say that the IA and OECD numbers are actually compatible. They measure different things, but you can add them up. IA focuses on developing emerging countries and price-driven subsidies. OECD measures other things, the fiscal cost in developed countries mostly. So it's not incompatible. And the two organisations actually work together. The price gap method of the IA is not the definition, it's a method. Do you have a definition that is actually broader and encompasses all sorts of budget transfers and tax breaks? Jehan didn't introduce himself, but that's Jehan Sauvage at OECD. And he's done a lot of the work in building the OECD's data set and working on the G20 stuff. So he knows what he's talking about. So do you want to go to some of the questions then? Do you want to start? Are there any more? Do we take any more? Alex, go for it. So I was going to say something about definitions, but I feel like we've got it covered. My counter question actually to your question is, doesn't matter if we have complete harmonisation agreement on definitions or can we proceed with trying to achieve reform without that. But the real question that I'm going to ask, since I think that's been covered already, is I'm always interested in, because Oil Change International would do some analytical work on subsidies as well as campaigning and advocacy work around subsidies. A couple of you on the panel I know and a couple of you I don't, but I'm very keen from the perspective of people who do a lot more research-focused and analytical work on the issue, is what's the upshot of your analysis for actually achieving change in the world in terms of advocacy and campaigning. And I know that's something that is the job of the campaigning community to some degree to figure out, but I'm always interested to hear the insights of people who are doing deep analytical and scholarly work on this about what's the missing piece if we don't have a lot of reform that's happened so far. From your perspective, from where you said, what do you think is missing? OK, let's go to the panel and we'll have time to come back and save your questions. It's coming back. So I think I'll take the first question, which was a very interesting question about the assumption I used. Just to be clear, I did not assume direct causation between the reforms being used especially for climate policies, but it was, because it's a theory, it was more of hypothesising because it has been put forward, for instance, that it can lead, maybe that's not the primary reason in most cases where you have the reforms for climate policy, but it has been identified as a tool. So the major assumption was, suppose it was used as a tool, what conditions would lead it to be successful or not. So there's a big trade-off because it's a general theory that looks across a number of countries, so in some cases it might hold, in some it might not, but I think what is most important maybe is now getting country-specific studies which give us now the causal reasons or the causal links exactly to that, but this was more of getting an approximation of if it was used as a climate policy instrument. So I think it doesn't have a strong direct link because it's very difficult to establish that unless you get very empirical country or sub-national case studies to put that up. So there are a couple of questions here. First, I see subsidy reform as beneficial for many reasons, one of which is that it helps align your price signals with concerns over climate change, but I also don't see it as the major reason to reform subsidies, it's an important benefit of it. In terms of the definitions, subsidies go along and continue on. So in the center you have government grants, it's a cash transfer from one party in the state to one private entity, there's really no definitional problems it's easier to measure. As you start going to the more complicated transfer mechanisms, it becomes more difficult because you need to have a baseline condition that you're comparing it to in order to assess the subsidies. Sometimes those baseline conditions will vary from country to country, but I certainly believe you can make important progress without full consensus on those. It's also quite important to recognize that probably half of the definitional problems people raise are politically driven that they are trying to basically inject into the process in order to slow things down because when you have information out there, it makes it harder for them to defend what they're getting and sometimes what they're getting is a lot. So I think definitionaly we're certainly workable in some of the margins in terms of tax expenditures across countries or benchmarking credit subsidies or indemnification and insurance, that's an area where there can be improvements and there's a lot of missing data, but I think we're doing okay. Well, I started my presentation with saying that I don't think the discussion of definition is useful. Moreover, subsidy definition is being discussed when people basically want to sidetrack the conversation as Doug said for political reasons and that's exactly what some countries are doing in G20 and that's why it's got stuck. But at the same time I think one of the major lessons learned is that reform happens at the national level, not necessarily at the global level, I mean if you want to play to the gallery for domestic purposes, for countries like US it may work like you're not the only one who's doing the reform, but in practice all reform always happens at the national level for reasons not related mostly to climate change, mostly related to budgetary constraints of course. And the other upshot of the research is basically that if we're looking at supply side mitigation measures, it's about the lock-in of high carbon assets over the long term and in this respect it doesn't matter if it's production subsidy or consumption subsidy, they should be viewed together and dealt with together and again the opponents of this approach, the opponents of subsidy reform would do exactly what the Romans did or the British Empire did so they will divide and start arguing about details. So I think for campaigning organisations it's important just to make a very strong blanket statements about all types of subsidies. Well I think that we have to think outside the box, it's not just that subsidies are important for climate change, are important for different sectors, affect the prices, affect the market, so in developing countries the climate change is not the only problem we have, so avoiding subsidies can help to other problems, to the economy, to a market, to prices, so maybe it's important to develop policies considering different aspects, to be more rentable for the people talking about poverty or markets and not just climate change. More questions? Raddak is behind you and then one at the top. Hi Raddak Stefanski from the University of St Andrews and the Oxford Centre for the Analysis of Resource Rich Economy, sorry for that. I just want to ask you maybe, you might know about this a little bit more, measuring fossil fuel subsidies is just one type, that's just one type of subsidy is there, some mechanism around the world in the United States that measures other types of subsidies and that might help us also understand whether subsidies like that are big potatoes or small potatoes, so sure we're subsidising fossil fuels a lot but maybe we're subsidising other stuff as well and so it would be interesting to compare. I'm Neil McCulloch from myself, independent consultant. I wanted to ask a quick question to Kennedy, I missed your definition of what fossil fuel subsidy reform actually entailed and so therefore it was difficult to interpret your odds ratio results but the most peculiar result from your odds ratio was that it seemed that the provision of overseas development assistance gave rise to a 90% reduction in the chance of a fossil fuel subsidy reform, maybe I misinterpreted your graph but that would be surprising and interesting and there must be a story behind that. I also had a second question for the whole panel which was really to do with what are the sort of circumstances that give rise to fossil fuel subsidy reform. I think it would be fair to say, I'm a development economist and I think it would be fair to say that climate change does not feature highly at all in the thinking of most fossil fuel subsidy reforms. I've lived for eight years in Indonesia and was heavily involved in the analysis around the Indonesian fossil fuel subsidy reform and I assure you that no one mentioned climate change as the reason for doing it. One of the reasons that these things get done is because of fiscal crisis. The initial discussions about fossil fuel subsidy reform in Indonesia happened under the SBY presidency when they were spending 30% of their budget on fossil fuel subsidies and they realised that that really wasn't sustainable. Other reasons might be because you wish to shift resources into the things that you like. For example, Jacobi's fossil fuel subsidy reforms are primarily because he needed a mechanism for funding his health card and education card reforms which are key components of his political platform. So I'm wondering whether you can give some hints as to the sorts of things that actually drive fossil fuel subsidy reform when it does actually happen. Hi, Franz Mattsner with NRDC in the US. I'm really intrigued by how we mix the campaign element of this. So as a campaigner in the US, just one observation is one of the reasons we don't make the progress is because subsidies make a great political weapon and so both sides use them and we mix and match our terminology of this. One person's subsidy is another person's incentive. So I do think that the definitional aspect of this matters to some degree because we've run into this problem a lot. It's a subsidy when it's going to something you don't like, in our case fossil fuels and it's an incentive when it's going to wind and solar. And we don't have a great way to distinguish those pieces and both sides like to use that as a political weapon but nobody really wants to do the forms because the scale of the money at stake that's being identified is usually too small, at least in the US, to actually pay for the thing that they want. So it becomes just a horse trading. So putting that together as an observation, I'd love to get some advice from the panelists on how we could distinguish incentives from subsidies and also find a way to, kind of as a community, have a bigger number to play with because the bigger the number, that's been the problem, the President, Obama identifies $4 billion and it sounds like a lot of money but it doesn't pay for health care reform, it doesn't pay for the big ticket items. So how can we make that number as big as possible and distinguish it from the good stuff? And one from Laurie and then we'll, oh sorry. And one down here. Okay, Laurie, Sheila, that's two ODI, and then one from down here and then we'll stop. We've got time. So Laurie from ODI, thank you very much for your interesting presentations. It has to be quick. Yes. So for me, fossil fuel subsidies both to production and consumption really raised questions about the distribution of the benefits and costs of fossil fuel production activities and also about who carries the risk of investing in these industries. And I think that in many countries there are still very big misperceptions around this. For example, in the UK, the government is now providing a net subsidy to the industry while a lot of the, or a big share of the population still thinks that it's a big source of government revenue. So how do we actually shift the narrative? How do we make sure that we get a wider understanding of the distribution of costs and benefits of fossil fuel production and consumption? Sorry, back to back ODI. Sheila Whitley, I'm with the Overseas Development Institute. I'm very quickly, I'm wondering if actually shooting myself in the foot as I've started to work almost exclusively on fossil fuel subsidy reform. If actually in the case of countries where climate is important that we may have leapfrogging over subsidy reform where it becomes much more of a focus on moratoriums, climate tests, other tactics to actually undercut fossil fuel production that might make the subsidy fights irrelevant in the context of climate. Did you, I misunderstood perhaps, did you consider among other things prices and interest rates? I didn't see interest rates. If not, do you think that it will change the calculation of the present value? And for sure, do you think that the reduction in subsidies, so-called subsidies will help in the reform, we see that perhaps the reduction of subsidies will be compensated with reduction in taxes and royalties? So go for it. Who wants to start? Do we need a mic down there? There's a lot of stuff out here, so I'm going to do my best to respond. I think there's some themes that I was hearing. There are measures for all types of subsidies, that was your question. Indeed, you want to do that. You want your fiscal system, you want your treasury, you want an integrated budget that's looking at the ways government is using its power to redistribute resources and you want to be able to measure it. So tax expenditures are credit stuff that cuts across sectors. It's not granular enough. You can't see which sector, you can't see which projects. That would be a huge improvement. But that relates to the big number issue, because subsidies to some other industries also have pretty negative effects, also contribute to climate change. And if you could get an integrated metric of the subsidies to energy plus energy intensive industries, it would be quite useful. I disagree with this whole question of differentiating incentives and subsidies. I think it's not a useful discussion when in solar gets subsidies. Those subsidies are put in place for a particular reason, just like the subsidies to oil and gas were put in place 100 years ago for a particular reason. Then everybody says, well, you're treating this sector differently than that sector. It's fine to just say, I believe this group should be subsidized and here are the reasons and then fight that battle. And I think you're going to end up in a better place than trying to pretend that somehow you're different. In terms of this moratorium and spending cuts, I think it's an interesting strategy. When you start winning all those battles, maybe you can step back on the subsidy front. One concern I have is that moratoria work best for publicly owned resources and lots of production is on private lands. I'm also concerned that if you have a state-owned energy company, it's going to be very difficult to get that state to cap down on your supply. So should be pursued, could be useful, but I would continue to work on both fronts. So in general fossil fuel subsidiary formers, of course it's not by itself a game changer, so it should be viewed as part of a suite of policies which also include mitigation measures for the losers of this reform, which include green economy transition, which include other sources of revenue for the government, not just the fossil fuel industry. So it's quite a big package of measures and the beauty of fossil fuel subsidiary formers is that it's self-sufficient in itself in the sense that if you implement it, you save money. So you don't need climate change, you don't need other arguments and that's what has worked in the past. So the narrative, this is rather to your question, the narrative about fossil fuel subsidies has been historically very strongly influenced by agricultural subsidies and the definition we use within the WTO has been designed with other industries in mind, so agricultural subsidies are not a very inspiring example, but we can possibly learn something from them and it's billions and billions of dollars as well. So with your question with interest rates, so there is a very wonky aspect of fossil fuel subsidies which is the concessional element of credit support and public finance. It's very difficult to measure. The right conceptual approach is how you can do it, but yes, companies, especially state-owned enterprises, very often get access to cheaper finance, and this is an area where we can hopefully do more research, bring more transparency. Of course this will increase the number and talking about increasing the number, I think the best way is to treat both production and consumption subsidies together, but there is a danger of increasing the number to the level where it stops making sense, and unfortunately this is some of the criticism of the IMF estimates that includes externalities into their definition. It's not like it's devoid of sense, but it's not money that you can reallocate for other purposes. So in this sense, getting a bigger number can shoot itself in the food, so we have to be careful about this. Well, considering that in Mexico, the company who produce oil and electricity are a national company, so when they give the subsidy, they are going to the subsidy to their own company. It's just a transfer of the... Well, they sell the input in a low price, so that's a big problem, and maybe the first impacts are going to pass when the electric reform begins and the activity is because the private companies, I'm not really sure that they agree to sell their input in low prices, so that will be very interesting to analyse the first oil prices that they sell to the other companies and the impact in electricity prices, so I think that's the most important impact. Just for Kennedy. Okay, so I think I'll respond to the very interesting question on how I defined a policy, a fossil fuel subsidy reform taking place. So my conceptualisation had two qualifications. So one, it looked at it as a national law or policy that is a national legal framework, and number two, from the database that I got it, all the laws that were there had a relation to climate change, whether implicitly or explicitly. That was qualification number one. Qualification number two was database has the laws and policies disaggregated in different segments whether energy demand, energy supply, red plus, et cetera. So I looked at the energy supply one. So for those two qualifications, we could get a strong approximation. Not a perfect fit, but a very strong approximation to the hypothesis. And secondly, the peculiar thing about ODA was very interesting because one of the very strong arguments in literature is that the reforms are a very cheap economic tool or rather low-cost economic tool like you remove the subsidies especially for developing countries, you remove the subsidy from oil and you invest in more pressing needs like education, health, et cetera. But if someone else is paying for that, why should you? Secondly, so that means when you get overseas development assistance to pay for health, to pay for education, there's little incentive to take that money from fossil fuel subsidies and put it there. And they're very much linked as well to the political cycles. So most of the time, and I think it was rightly said so, the reforms are not especially for climate change related things, but they're more political in nature. For example, the politicians will have low-cost energy which is better for them especially for their election prospects. So for me that was my tentative argument, but I'm very much open to better arguments as to that peculiar finding. Thank you. Thank you very much, Kennedy. It is an academic conference and part of that is about peer review and learning. So speak to the panellists and talk to them about their papers and help share ideas. I'm just going to say very quickly to wrap this up that yesterday we found it really hard to get to the pub that you'd organised for us to go and have a drink at before this conference because they were filming transformers. And they were filming it here in Oxford and when I was a kid it was robots in disguise, transformers, robots in disguise. Now, production subsidies are a bit like those transformer robots. If you can find them and if you can identify them and you don't really know their impact from looking at them on the outside you really have to sort of get under the hub to figure them out and to work out what reserves they unlock upstream and how to undo them. So like transformers we also know we've heard here today there are goodies and there are baddies as well. And some producers' subsidies are much worse than others especially the ones up front that Yvette talked about. But not all producers' subsidies are bad and you know this is what you raised as well about subsidies for solar and renewables and so on. So that's exactly what our panellists have helped us to understand today. Get to understand these transformer producer subsidies they start right at the beginning of the energy chain and help us to understand how we can beat them and how we can change them. So let's really thank the panellists in a traditional manner.