Julian Alston, University of California, Part 1 – ‘Driving productivity growth through R&D…’
So, I’m going to talk about a few of my hobby horses today. First, rates of return to agricultural R&D. I think even the sceptics have accepted now that agricultural R&D does generate high rates of return – very high rates of return. They don’t always reach the obvious conclusion from that; which is that we’re under investing – and so that’s something to carry around with you, that if you buy that the rate of return to agricultural research is high, as we say it is, that means we’re under investing. So, um … that’s one thing. Another one is what’s happening to spending on agricultural R&D – so in spite of what I just said about what rates of return imply – that we should be spending more – we’re going in the wrong direction. So, in the high income countries in particular there’s been a slowdown in the rate of growth of spending on agricultural research and that money is being spent on different things. So, a diminishing share of the total is being spent on farm productivity growth enhancement. That’s true – particularly in the more developed countries. Middle income countries – particularly Brazil, China and India – it’s a different story. And then, er, productivity. We’ve heard quite a bit about this today and I’m glad that more and more people are saying this now. I’ve been saying it for a long time and it used to be hard to get people to listen but it’s become quite clear to many people now that there’s been a substantial slowdown in the rate of growth of agricultural productivity. You’ll still get people disagreeing about that – it’s hard to measure, hard to be precise about it but I think the picture is becoming clearer. And again, it’s different in different parts of the world. So in China, India and Brazil – not as much of a slowdown as in the rich countries. Maybe it’s a coincidence that those are the countries that are backing off from agricultural R&D support. So, what are the implications? Well, I think we need to do something differently if we’re going to reinvigorate spending on agricultural R&D – particularly in the public sector. If we don’t do that there are some bad consequences in terms of long term. Agricultural R&D takes a long time to effect productivity – then it affects it for a long time. The long term consequences of not doing enough are … the world food supply and demand balance is different. We have more people poor and hungry in the world, we have more strife related to that, we have more pressure on the natural resource base. If we’re the country that’s relatively slow we lose competitive capacity. Okay, that’s my story. Er, this is, er a summary of a meta-analysis we did about ten years ago – a bit longer – summarising all the evidence in the literature on rates of return to research up until then. There’s a lot of studies. Nearly 300 studies. Some of them estimated more than one rate of reutrn, so we had some 1800 estimates of rates of return – and the average of that was 82% per annum real. That’s a pretty good rate of return. Now we think a lot of these estimates are probably biased up but the answer is compelling – and the mass of that distribution lies between about 20-40% real rate of return. For people who invest, you’d be happy with that, I think. New evidence – this is a bit vainglorious, I’m sorry – but this is the best work in the area. I say that because it’s my work … This book is called ‘Persistence Pays’ – and it’s for a reason. It took us 15 years working a fair bit of the time on this to generate it. It’s about the US, er, and interesting perhaps to people who aren’t interested in the US because we have relatively good data in the US – so it’s based analysis of over 100 years of detailed, state level data on agricultural research and extension spending and the consequences of that for agricultural productivity in the 48 contiguous states, measured using really wonderful data over the post-WW2 period. So we have lessons from that for the rest of the world. We identified challenges in here, some of which Terry brought up in terms of modelling spill overs. Not the spill overs that people sometimes talk about with respect to environmental externalities or to other sectors. It’s within the sector but among spaces. So, when Victoria does research it generates benefits in Tasmania or New Zealand – and that affects Victoria’s incentives to do research and we have to account for that in modelling policy and measuring the consequences of policy. Lags are really long. In our research we had much better data, much longer series than most people. We could measure the lags – and they’re much longer than most people think or most people have used in their models; and that matters for the measures. And then there’s a ‘role of maintainence’ research that’s special for agricultural R&D. Er, so in, er … in ecology they talk about the red queen effect; which is the idea that when species co-evolve you’ve got to work really hard just to break even with all the other species that are competing with you – and it applies in agriculture, so if we stop doing research productivity wouldn’t just stay where it is it would fall because pests and diseases get better. So, when we’re doing innovation in agriculture we look at average yields and so forth and we’ve got to deal with the fact that the location of production is changing, so we have to do adaptive research just to hold productivity constant in a new place. And because pests and diseases are evolving we have to do maintainence research and um, just to keep up with them. It maybe half or two thirds of the total research spending is maintainence research – that’s necessary just to hold productivity where it was Okay … So, we did this work on the United States and this is summarising … this is like the bottom line from the book. Looking at the first column you see across the 48 states the average benefit/cost ratio was 21:1. That means – you put down a dollar today you get eventually benefits that are equivalent to $21 today. That’s a really good return. I like that kind of investment. In addition to the $21 at home you generate another $11 in other states. So, the next number is 32.1:1 – that’s the national benefit/cost ratio. That’s really big. And there’s a range across the states from ten to seventy to one. USDA research generates about 18:1. That’s a really big benefit/cost ration. But we’ve done a better job than other people of dealing with all these awkward measurement problems – so even though our benefit/cost ratio seems really big, our estimate of the internal rate of return is only around 20% per annum – down the low end of that middle So these are really wonderful investments returning benefit/cost ratios of twenty, thirty to one with an internal rate of return of 20%. Pretty good investments. Okay, this is total science spending in the United States. I’m using the US as a metaphor because we know a lot about it. So, they’re spending nearly $400billion in 2008. Agricultural science around 2% of that. You compare the food and agriculture sector versus total science and there’s a difference in how it’s done. In the rest of the business, business has a lot to do with it – whereas in food & agriculture there’s a much bigger share of private sector in total. That’s because the nature of market failures in agriculture is different – because of the structure of the industry. This is spending in the United States, 1950-2009, and you can see that it flattens out a bit, just like the Australian story. Um, private and public doing roughly equal amounts recently. In the private sector quite a big share of that money is being spent on food processing. It’s beyond the farm gate. We’ve got to keep that in mind. So, I’m concerned about farm productivity – and the private sector is doing some of that but it’s doing a lot of stuff that isn’t about that. This is looking at the growth rates of spending and we can there’s been a slowdown in growth of spending – as Terry was talking about here. What’s the money being spent on? So this is State Agricultural Experiment Station research. Two thirds or three quarrters of the total is done in State Agricultural Experiment Stations, like the one where I work in California. the rest done by the Federal Government in USDA labs. A lot of the money coming into the State Agricultural Experiment Stations comes from the USDA or other federal sources. What this picture shows is that at the tail you see the federal share of total spending in the State Experiment Stations – going up since about 1980 But the USDA share of that is going down. So, what’s happening is where I work we’re getting a bigger share of our money from the Federal Government but it’s not from USDA. It’s from National Science Foundation, Department of Energy National Institutes of Health. They’re not interested in funding farming research – so the farm productivity share of State Agricultural Experiment Station research, er … this has got a crooked scale on this … it starts at 55% not zero, okay. It hasn’t fallen as far as it looks. So, it was around 70% in 1984 and is down below 57% now. So, it’s gone down. We’ve got this for every state So distribution? The top one’s 1976 – had a mean of 65% on-farm and that mean has shifted down to 56%. The whole distribution has shifted to the left – and that is because there’s much more interest in funding research into food safety, health & nutrition, environmental issues – things like that rather than farm productivity Um, … the world as a whole … agricutural science is, er … about 5% of the global science budget – and we can see that it’s distributed around the world. Quite a big share comes from a small number of countries. The United States does roughly 20% of the total – roughly half and half private and public investments in the United States. This is a measure of intensity ratios, like what Terry was talking about – among different countries. If you look on the left hand picture you can see that, er … and these are showing over time, going from the 70s, 80s, 90s, 2000s. So this intensity – research spending per dollar of output has gone up from, in the high countries, from one to three percent. In the low and middle income countries – stay at around 0.5%. There’s quite a bit of variation among countries going up over time but noticeably in the rich countries is where the private sector activity is. So that matters if you care about people going hungry in the rest of the world. Looking at growth in rates of this speding – and we can see a slowdown in growth of spending everywhere really but, er … especially in the low and middle income countries. Now, last topic. Productivity. Another book from last year, er … this one … that first book you can buy It’s very expensive. I don’t get significant royalties from it so I don’t care if you buy it. This one’s free so you can help yourself, get it off the web. This has got chapters on many countries talking about changing agricultural productivity patterns in the world. We wrote this book because we became concerned about this issue and wanted to find out whether what we were observing in the US and a couple of other places was more pervasive. If you don’t like reading the whole book – even though it is free – you don’t have to read the whole thing, you can get a summary in Choices magazine, where we picked out the main messages and had short articles on the main places where there were stories to tell. The main points of this is that there’s evidence of a significant and pervasive slowdown in agricultural productivity growth since about 1990. And er … there are some exceptions. China is a really important exception, with faster growth. The former Soviet Union and the central European countries are the opposite – and we think in both cases institutional reform had a part to play in that pattern. We used lots of different measures. We used crop yields – which is kind of partial productivity measure – production per unit of land or labour and in a small number of places we have good multi-factor productivity measures, like what Terry was showing us. We have them for Australia, Canada, the US and the UK. They all show the same general story. There’s a consistency between the patterns. In a place like the US where multi-factor productivity growth has slowed down a bit, crop yield growth has slowed quite a bit too.