Green innovation

Philippe Aghion interviewed by Romesh Vaitilingam, 19 November 2010

Unfortunately the file could not be found.

Open in a pop-up window Open in a pop-up window


Download MP3 File (6.84 MB)



Topics:  Environment Industrial organisation

Tags:  climate change, clean energy

Romesh Vaitilingam interviews Professor Philippe Aghion for Vox

November 2010

Transcription of an VoxEU audio interview []

Romesh Vaitilingam: Welcome to Vox Talks. A series of audio interviews with leading economists from around the world. My name is Romesh Vaitilingam and today's interview is with Professor Philippe Aghion from Harvard University. Philippe and I met at the European Economic Association's annual meeting in Glasgow in August 2010. Where we spoke about his research on what governments can do to encourage the private sector to do more green innovation.

Philippe Aghion: Mostly the economists working on climate change have mainly looked at the idea that we should reduce production technologies that use dirty inputs by means of introducing carbon tax of carbon trade, et cetera. But largely disregarded has been the fact that people in that firms not only produce but they also innovate. And they may decide to innovate in technologies that are dirty or technologies that are clean. In technologies, for example, that make more use of fossil fuels or technologies that allow you to save on fossil fuels or to use alternative sources of energy. And we know that to overcome the climate change problem or to mitigate the consequences of climate change, innovation and the diffusion of technologies, of green technologies, will be absolutely important.

Having worked on innovations and growth for a while now, I thought that it would be very natural to introduce endogenous technical change in models of environment and growth.

Romesh: Can you give us a general feeling of what we know about public interventions in order to have an impact on innovation, as economists? What do we understand by how those methods work?

Philippe: We know that there are various ways. You can, for example, carbon tax on its own will tend to direct innovation towards clean technologies. And in fact this morning I presented a paper where we use data from the automobile industry - cross country panel data - which shows that indeed whenever you have an increase in the price of fuel or in the increase in the fuel tax, that leads firms to direct innovation activities towards cleaner. But there are other instruments which besides carbon tax, which are direct subsidies to clean R&D or to the diffusion of clean technologies. And in fact, the optimal portfolio of policies is not just carbon tax but a mixture between the carbon tax and these other policies. So that not all the burden is on the carbon tax. And of course, factoring in technical change leads you to think deeper about how to conceive of this optimal portfolio of policies.

Romesh: Can you explain a little bit about how you see these interventions operating. How they work in terms of encouraging firms to invest more in clean innovation as opposed to dirty innovation?

Philippe: Yeah. They encourage more because if you put a carbon tax you reduce the profits that would be earned by using fuel inputs. But we know that the profits are what rewards innovations. So if you reduce profits from dirty activities you also reduce the reward that would accrue to an innovator that innovates in dirty technologies. And therefore that in turn will tilt his incentives towards innovating in clean innovations, you see? That's how carbon tax works, but of course the direct subsidy to clean innovation will also have the effect of shifting R&D effort towards clean innovation. The question is to which extent those substitute or compliments. Of course, the advantage of subsidies is that you deal directly with incentive and you don't put too much pressure on current consumers.

But the advantage of the carbon tax is that carbon tax is non targeted. In a sense that you don't say, you don't target R&D subsidies to particular activity. So very often we're worried about industrial policies whereby you would pick winners for example. So it's true the R&D subsidy-targeted might have more of a flavor of industrial policy. But on the other hand just using the carbon tax also has its downsides.

Romesh: I'd like to come back to the issue of industrial policy in a moment. But I'd like to talk a little bit about the data you put up. You put up some interesting data on citations patents.

Philippe: Yes.

Romesh: And patent citations as kind of an indicator of R&D changes over time. Can you give a flavor of the story you told there?

Philippe: So we had a data set, it's a world statistical data. The patents are on the web, but those all concerned firms that have patented in Europe. So what happened, you have the European Patent Office. For any patent at the European Patent Office they look also if the same thing has been patented elsewhere. So they have this way a record of patents not only in Europe but also they can find out if those innovations that have been patented at the EPO - the European Patent Office - have also been patented at the Japanese Patent Office or at the US Patent Office. And also the data set we have is very interesting because we know who has patented. So we know the history of patenting. So we know for example if those who innovate dirty today are those who have innovated dirty more in the past. And in fact what we established in the presentation, in the paper I presented this morning, is that there is what we call past dependence in the direction of technical progress.
If you have innovated dirty a lot in the past, you tend to innovate dirty more today. You tend to continue to do things you were good at. There is no reason a priori where it should be the case. You could have argued that there are decreasing returns. You could have argued that because I have innovated a lot dirty in the past, I run into decreasing returns, and I rather start innovating clean.

But in fact that's not what happens. If you're a good cook and a bad dancer and you've innovated in cookery cuisine in the past. Well, you prefer to innovate in cuisine now than start innovating in dancing. You see what I mean? And that is very much the way it works. And it has big implications this past dependence because typically firms in our economies have innovated a lot in dirty technologies in the past.

And therefore if you have no intervention from the state, they will keep on innovating dirty today. Which means it will be much, much harder over time to make them produce and innovate in clean technologies. You see what I mean? So it's very important, this past dependence because it implies that there is urgency to intervene. If you don't intervene today firms, will keep innovating dirty. So the technological gap between dirty and clean technologies will increase. And it will be much more costly to restore the balance in the future, to make firms produce and innovate clean in the future.

Romesh: So intervention is in a way saying, "Give up your cookery because it's damaging everybody, and we're going to shift you over to dancing. We're going to give you an incentive to change over."

Philippe: Exactly, exactly.

Romesh: In the data do you see differences across countries in terms of the balance between clean and dirty innovation?

Philippe: We know that for example in Japan or in Denmark…they have done more in terms of clean innovations in general. Not just in the automobile industry. US has done very little for example. France not too bad. China not too bad actually. China is taking seriously clean innovation. Sometimes people believe that China is not playing the game at all of the green economy and in fact they are. They are getting into green innovation. So what is particularly striking is the fact that the US has been very much out of it.

Romesh: But you have seen an increase over time, generally, across the world in clean innovation. What is this a result of? Do you see this as a result of previous interventions that governments have made, changes in fuel prices for example?

Philippe: I think changes in fuel prices already had a bit of an effect on inducing it, and I think people now foresee that it will be more and more important to go towards clean cars, etc. I think it's been because of the increase in the price of oil, the fear that we might not have access to oil so easy. Partly as a result of carbon tax in some countries. What's interesting is that in spite of a lack of public policy in this, private firms are anticipating things will happen. It's very interesting if you look at the private markets, you can see that there is an interest for green technologies from the private sector. If you look at the venture capital there is an action in the clean sector, actually.

Romesh: There's also, of course, a lot of interest in governments in encouraging green technology. You've mentioned some ways they might do it, but there seems to be almost a focus much more these days on industrial policies, governments thinking we can pick winners in green technology and we can push it. Do you still go along with the traditional economist view that industrial policies are a very dangerous route to go?

Philippe: It's true that I think was good to be cautious. To pick winners is not a good thing. It's not a good thing because the government might not know who the winner is because it opens the door to corruption and lobbying, so it has a lot of downsides. I think it was good that economists in the past 20 years have pointed to the downsides of the top down pick winner policy. Does it mean you should have no sectoral policy? I believe not because if you believe in particular, and the climate is a case in point, we know that under laissez faire if you don't intervene, firms will innovate dirty. So you have to step in and say we will induce you to innovate clean. 

Now, that doesn't mean that you pick one winner, but that means that you will encourage firms that want to pursue production and innovation in clean technologies, but the idea would be to say whoever enters and does that, I could subsidize or I could help. So it would be an open subsidy, it would not be for one firm only, so it would be competition friendly. It might even spur competition because, maybe, firms would have differentiated from each other and by focusing on sub sectors, you push firms to go into the same sector sector and therefore compete more closely in that sector. The idea that sectoral policy will always be the opposite of competition policy, that I think is going too far.

One thing is to pick one winner firm, but one thing is to say I have a sectoral policy which is competition friendly and which is open to anyone who goes into the sector. A good argument for that is there are externalities like past dependence externalities that are not internalized. Another argument for industrial policy is also that capital, physical and human capital, does not always relocate on its own because you have your credit constrains. That's another reason why government might want to step in. Again, it should do it in a way which is competition friendly.

Romesh: What should be the role, do you think, in public investment in green R&D? We know that public sector investment in science is very important for private sector to pick it up and use it and for growth. What should be the role in this particular area of green innovation, do you think?

Philippe: I think there is a synergy between public and private intervention. I think the public innovation can convene together with private investment. I very much see intervention in co financing arrangements, for example. It's good to co finance because if it turns out the project is bad, the private sector withdraws and then the public sector knows it has to withdraw as well. You see, that's one way also to make sure that if you pick the wrong winner you don't pick it for too long. I think there are ways to do it which would involve partnership between public and private. My feeling is that once, at some point, because of past dependence, once clean technologies have become very sufficiently advanced, past dependence would play then in reverse.

It would guarantee that private sector firms under laissez faire would keep innovating clean. The idea is that the government comes in, steps in to redirect attention to clean, but after a while there is no need for the government to stay because the market on its own, thanks to past dependence, will make clean firms innovate clean in the future. That's why you can see this view of the world is really a kind of third way view of it, where you have partnership between public and private sector.

Romesh: Philippe, you spoke in your talk about the whole area of the microeconometrics of climate change being almost virgin territory. Can you elaborate a little bit more on that?

Philippe: Yeah. So far what people have done is they have used calibration to calibrate macro models, so there is the Nordhaus model. We saw this morning elaborations on the Nordhaus model. What people do usually is they calibrate at the micro level. Let's say we want to map the share of capital and labor that we observe in practice, we want to map with the average growth rate that we observe in practice, we want to map with how concentration of CO2 translates into temperature increase at the micro level, so we do micro calibrations of models, which is already a very good first step.

I think there are important parameters, such as elasticity of substitution between clean and dirty inputs or the productivity of clean and dirty innovation or clean and dirty imitation of technology. Those parameters should be estimated, not just calibrated, should be estimated using microeconometric analysis. I think that's the way to go. What we tried to do this morning was to test the past dependence assumption using micro data.

I think in a very similar vein, we should use this kind of data to estimate these parameters and then we can recalibrate. You see what I mean? But I think it's really important now to replace calibration, or at least to add to calibration exercise, microeconometric analysis. We put IO and microeconometrics into climate change and growth. I think that's the kind of big agenda for the future.

Romesh: So final question, where does your own research fit into this? Where do you see your development in this program?

Philippe: I think it's very much in line with what I've tried to do. I don't know if I brought anything to economics, but one thing I tried to bring to economics was to put IO into growth. Industrial organization is at the heart of the growth process and you should understand market structure, the organization of firms, all these things are important to understand the growth process. I try to do the same here with climate. You need to go into the details of how firms operate, how they mix clean and dirty, what kind of innovation technology they have access to. You need to go into those details to get back into growth. It's part of this agenda of putting industrial organization in growth.

Romesh: Philippe Aghion, thank you very much.

Philippe: Thanks very much.


Professor, College de France and London School of Economics; CEPR Research Fellow


CEPR Policy Research