Abandoning a Paper

Now and then it's time to give up on a project. In September 2018, I attended a climate econometrics conference at Frascati near Rome. For my presentation, I did some research on the performance of different econometric estimators of the equilibrium climate sensitivity (ECS) including the multicointegrating vector autoregression (MVAR) that we used in our paper in the Journal of Econometrics. The paper included estimates using historical time series observations (from 1850 to 2014), a Monte Carlo analysis, estimates using output of 16 Global Circulation Models (GCMs), and a meta-analysis of the GCM results.

The historical results, which are mostly also in the Journal of Econometrics paper, appear to show that taking energy balance into account increases the estimated climate sensitivity. By energy balance, we mean that if there is disequilibrium between radiative forcing and surface temperature the ocean must be heating or cooling. Surface temperature is in equilibrium with ocean heat, and in fact follows ocean heat much more closely than it follows radiative forcing. Not taking this into account results in omitted variables bias. Multicointegrating estimators model this flow and stock equilibirum. The residuals from a cointegrating relationship between the temperature and radiative forcing flows are accumulated into a heat stock, which in turn cointegrates with surface temperature. If we have actual observations on ocean heat content or radiative imbalances we can use them. But available time series are much shorter than those for surface temperature or radiative forcing. The results also suggested that using a longer time series increases the estimated climate sensitivity.

The Monte Carlo analysis was supposed to investigate these hypotheses more formally. I used the estimated MVAR as the model of the climate system and simulated the radiative forcing series as a random walk. I made 2000 different random walks and estimated the climate sensitivity with each of the estimators. This showed that, not surprisingly, the MVAR was an unbiased estimator. The other estimators were biased using a random walk of just 165 periods. But when I used a 1000 year series all estimators were unbiased. In other words, they were all consistent estimators of the ECS. This makes sense, because in the end equilibrium is reached between forcing and surface temperature. But it takes a long time.

Each of the GCMs I used has an estimated ECS ("reported ECS") from an experiment where carbon dioxide is suddenly increased fourfold. I was using data from a historical simulation of each GCM, which uses the estimated historical forcings over the period 1850 to 2014. A major problem in this analysis is that the modelling teams do not report the forcing that they used. This is because the global forcing that results from applying aerosols etc depends on the model and the simulation run. So, I used the same forcing series that we used to estimate our historical models. This isn't unprecedented, Marvel et al. (2018) do the same.

In general, the estimated ECS were biased down relative to the reported ECS for the GCMs, but again, the estimators that took energy balance into account seemed to do better. In an meta-analysis of the results, I compared how much the reported radiative imbalance (=ocean heat uptake roughly) from each GCM increased to how much the energy balance equation said it should increase using the reported temperature series, reported ECS, and my radiative forcing series. A regression analysis showed, that where the two matched, the estimators that took energy balance into account were unbiased, while those that did not match, under-estimated the ECS.

These results seemed pretty nice and I submitted the paper for publication. Earlier this year, I got a revise and resubmit. But when I finally got around to working on the paper post-lockdown and post-teaching things began to fall apart.

First, I came across the Forster method of estimating the radiative forcing in GCMs. This uses the energy balance equation:

where F is radiative forcing, T is surface temperature, and N is radiative imbalance. Lambda is the feedback parameter. ECS is inversely proportional to it. The deltas indicate the change since some baseline period. Then, if we know N and T, both of which are provided in GCM results, we can find F! So, I used this to get the forcing specific to each GCM. The results actually looked nicer than in the originally submitted paper. These are the results for the MVAR for 15 CMIP5 GCMs:

The rising line is a 45 degree line, which marks equality between reported and estimated ECSs. The multicointegrating estimators were still better than the other estimators. But there wasn't any systematic variation in the degree of underestimation that would allow us to use a meta-analysis to derive an adjusted estimate of the ECS.

This is still OK. But then I read and re-read more research on under-estimation of the ECS from historical observations. The recent consensus is that estimates from recent historical data will inevitably under-estimate the ECS because feedbacks change from the early stages after an increase in forcing to the latter stages as a new equilibrium is reached. The effective climate sensitivity is lower at first and greater later.

OK, even if we have to give up on estimating the long-run ECS, my estimates are estimates of the historical sensitivity. Aren't they? The problem is that I used the long-run ECS to derive the forcing from the energy balance equation. So, the forcing I derived is wrong. It is too low. I could go back to using the forcing I used previously, I guess. But now I don't believe the meta-analysis of that data is meaningful. So, I have a bunch of estimates using the wrong forcing with no way to further analyse them.

I also revisited the Monte Carlo analysis. By the way I had an on-and-off again coauthor through this research. He helped me a lot with understanding how to analyse the data. But he didn't like my overly bullish conclusions on the submitted paper and so withdrew his name from it. But he was maybe going to get back on the revised submission. He thought that the existing analysis which used an MVAR to produce the simulated data was maybe biased unfairly in favour of the MVAR. So, I came up with a new data-generating process. Instead of starting with a forcing series I would start with the heat content series. From that I would derive temperature, which needs to be in equilibrium with heat content and then using the energy balance equation derive the forcing. To model the heat content I fitted a unit root autoregressive model (stochastic trend) to the heat content reported from the Community GCM with the addition of a volcanic forcing explanatory variable. The stochastic trend represents anthropogenic forcing. The Community GCM is one of the 15 GCMs I was using and it has temperature and heat content series that look a lot like the observations. I then fitted a stationary autoregressive model for temperature with the addition of the heat content as an explanatory variable. The simulated model used normally distributed shocks with the same variance as these fitted models and volcanic shocks.

As an aside, the volcanic shocks were produced by the model:

where rangamma(0.05) are random numbers drawn from a standard gamma distribution with shape parameter 0.05. This is supposed to produce the stratospheric sulfur radiative forcing, which decays over a few years following an eruption. Here is an example realisation:

The dotted line is historical volcanic forcing and the solid line a simulated forcing. My coauthor said it looked "awesome".

So, again, I produced two sets of 2000 datasets. One with a sample size of 165 and one with a sample size of 1000. Now, even in the smaller sample, all four estimators I was testing produced essentially identical and unbiased results! I ran this yesterday. So, our Monte Carlo result disappears. I can't see anything unreasonable about this data generating process, which produces completely different results to the one in the submitted paper. So, I don't see anything to justify one over the other. So, this was the point where I gave up on this project.

My coauthor, who is based in Europe, is on vacation. Maybe he'll see a way to save it when he comes back, but I am sceptical.

Robots, Artificial Intelligence, and the Future of Work

I think that robots/artificial intelligence and the future of work is a hugely important topic. This is a very active research field but it seems to me that people (some informed by this research but most without referring to the research) are rushing to one of two conclusions. The first of these conclusions is that up till now economic growth has resulted in rising wages and full employment and so it surely will in the future too. The other is that robots must mean structural unemployment and so the solution is to introduce universal basic income or some similar redistributive policy.

I don't think either is necessarily true. In the past, the elasticity of substitution between labor and capital seems to have been less than one - both inputs were essential in production. Also, the two inputs are q-complements - an increase in capital per worker results in an increase in the marginal product and, therefore, wage of a worker. But it's possible that now, or in the future, that the elasticity of substitution between labor and capital is or will be greater than unity so that labor is not an essential input. Or that there are techniques that are designed just to use machines. Acemoglu and Restrepo (2016) assume that as some low-skilled tasks become automated other new high-skilled tasks are introduced. But there may be limits to people's cognitive ability. Most people aren't intelligent enough to be engineers and scientists. And the people that are intelligent enough now, might be worse than artificial intelligences in the future.

The other premature conclusion is that, definitely things are different now, and robots will result in structural unemployment or immiserizing growth, so that government intervention is needed. Usually, universal basic income is mentioned. Sachs et al. (2015) argue immiserizing growth is possible. This is one of the better papers out there I think, but still the framework is quite limited. as technological change is exogenous. It is possible that there is some self-correcting mechanism similar to that in Acemoglu's (2003) paper on capital- and labor-augmenting technical change. In that model, capital-augmenting technical change is possible for a while, but it introduces forces that return the economy to a pure labor-augmenting technical change path. Another important question is whether people have a preference to have at least some of the goods and services they consume produced by humans. Sachs et al. assume that utility is a Cobb-Douglas function of automatable and non-automatable goods. That means that consumption of human made goods could become infinitesimally small in theory.

I think we need to consider a range of models as well as empirical evidence before we can say what kind of policy, if any, is needed.

I wanted to do research on this and started doing some research on this but concluded that it is not realistic given my limited time for research because of my administrative - I am still director of our International and Development Economics Program - and parenting roles and my existing research commitments. Twitter isn't the only reason that I haven't updated this blog in 2 1/2 months. Comparative advantage suggests to me that I remain focused on energy economics. However, I think that the Crawford School of Public Policy should be looking at these kind of issues, and I am trying to encourage that. This is going to be one of the key policy questions going forward I think.

Article Published Today in PLoS ONE

My paper "High-Ranked Social Science Journal Articles Can Be Identified from Early Citation Information" was published today in PLoS ONE. For background on the paper, check out my blogpost or this story on the Crawford School website.

I'm travelling tomorrow to Kassel where I will be working with Stephan Bruns on what we think is a really awesome (Stephan's words :)) extension to this paper. We are also hoping to finish an econometric theory paper we are writing.

Checking In

I haven't been blogging much lately - things have been very busy (started teaching, trips etc.) so I haven't had time to blog on papers I have read, policy issues and the like and several projects are near completion but not quite there and so I neither have anything to report on them nor any preliminary literature review etc. I can put up as blogposts. That pretty much covers the sources of content for this blog. This should change over the next month or so as some of these projects are finalized.

There is a little news to report. My paper on "Energy and Economic Growth: The Stylized Facts" (one of the almost complete papers) was accepted for the IAEE conference in New York City. So, I expect I will go to that meeting in mid-June. We got another revise and resubmit on the traditional and modern energy paper. Another, because we already had an R&R from another journal that then rejected our revised version. The AARES conference this month was a lot of fun. It looks like some of the papers will be re-presented here in Canberra for AARES members that couldn't make it to the conference. I'll let you know when my paper is scheduled. I also will present a paper on the same topic (emissions and growth) as a seminar at the Arndt-Corden Department of Economics here at ANU on 1 April. Maybe there is a reason why that date was still free :)

Last Thursday and Friday I had a research meeting with Jack Pezzey and Astrid Kander. We discussed our work on modelling the Industrial Revolution. I think we have a viable strategy for overcoming this setback. Location: Coogee Beach. Astrid has been visiting Sydney for this month working with researchers at the University of New South Wales and so Coogee was the perfect place for her to stay. Certainly, a great place for a meeting :)

Cuts to ARC Funding, Strategic Research Priorities

In yesterday's mid-year budget update, the Australian government announced cuts to ARC funding. $103 million will be cut over a 4 year period and diverted instead to medical research. This actually amounts to about 3% per year of the ARC budget. So, it's not as bad as one might think at first.

Back in June the government announced new Strategic Research Priorities to replace the previous National Research Priorities. I heard that these may be undergoing some revision. We'll only get new funding rules and instructions for all schemes in January. But I think we should assume that we will need to address these new priorities in ARC proposals to be submitted in the upcoming rounds at the beginning of 2014.

Information for Prospective PhD Students

In the US, a proposal is not a requirement for applying to do a PhD. Usually, students do one to two years of coursework before writing their proposal and it is enough to confirm that a department has faculty specializing in the general area that a student is interested in, say energy economics or climate change in my case. But, here in Australia, we expect students to submit a detailed proposal, even though in practice this proposal will be extensively revised after they start study. And in our economics program there is a year of coursework before students switch to research only. It makes sense to me for a student to select an area of interest to the potential supervisor and then discuss with the supervisor how to develop the proposal. Of course, a student might have a burning issue that they can't wait to conduct research on. But I doubt there are really many such cases. I certainly didn't know what I wanted to research. This was one of the main reasons I went to study in the US rather than remain in the UK. In the UK I would have needed to submit a proposal with my application.

Jack Pezzey has some good guidelines to help potential PhD students think about the application process in Australia. He recommends developing the proposal in consultation with him. He also has a great summary of his current research interests. That was meant to be one of the purposes of this blog. I used to have a research page on my website. But I scrapped that when I started this blog. I've decided it's time to put up a research page again. I hope this will be useful to give potential PhD students and collaborators an idea of what I am currently working on or would be interested in working.

Some previous thoughts on PhD applications.

Visitors

We just hosted a quick visit by Bob Costanza and Ida Kubiszewski to the Crawford School. Bob gave a public lecture and Bob and Ida participated in roundtable discussions with PhD students and faculty from the Crawford School and the Fenner School.

I have some more colleagues lined up to visit later this year. Next month, Stephan Bruns will be visiting Crawford for the whole of June. He will be working with me on our energy-GDP causality research. He will give a presentation in our economics PhD seminar series. At the end of the month, Christian Gross, also from Jena, will visit briefly and give a seminar. In September/October, Astrid Kander will visit to work on our ARC project.

The "Google Way" to Visualize Research

Yesterday, I showed one potential way of visualizing my research areas. Another way is using the web to tell us just like Google does. In fact, I did a Google Scholar search on myself and then put the results (with some cleaning up) into Wordl:



I could have also cleaned out my name from the data first (using the replace function in Word) but decided against that. Some of the topics from yesterday's post are clearly here but you need to know how to "connect the dots" to understand the links between them.

Visualizing Research Areas and Links

I've been trying to give an overview of my research for a presentation and came up with this:



I've tried to fit various of the topics I've worked on in the last decade within three disciplinary fields and the overlaps between them. In the middle I put "Meta" to represent my interest in meta-analysis and bibliometric analysis. This doesn't cover everything. My recent energy efficiency stuff is subsumed into energy and growth or EKC but it mostly covers the relevant topics. The variety of topics is one reason that I called this blog "stochastic trend". It's something of a random walk around these various areas of interest.

BTW the image was created in Powerpoint 2004 (Mac version) using "Autoshapes". I set the fills of the circles to 50% transparency to get the mixed colors where the circles overlap. There is a special tool for Venn Diagrams in the most recent version of Powerpoint.

Powerpoint Slide for Crawford School Research Retreat

Call for References: The Role of Energy in Economic Growth

Back in April I did a series of posts on energy and economic growth which serialised a review paper I was editing on the topic. A couple of those have been among my post popular posts in the last few months. Now I am revising the paper again for the final version. The referees made a lot of comments but I am also reading some additional papers and seeing if they fit into the story. So if you think there is something I should include please contact me and I'll be happy to consider it and if I use it include you in the acknowledgments. I already have 175 references though and the referees complained about some things I hadn't included so I'm not planning on adding a lot of stuff. But if it is important I will. I will put out a working paper when this revision is complete. Use the search function in Blogger to see if your reference was already included in one of the series.

Final Hub Report

The final report for my Hub project titled: "Modeling International Trends in Energy Efficiency and Carbon Emissions" is finally up on the EERH website. I've highlighted the key results in previous posts so not so much to add here. This paper also has a lot of literature review and a theoretical model as well as the econometric model and results. Originally I had more radical ideas of how to use the theoretical model but for various reasons decided to abandon them. The main reason was that I decided to incorporate the variables that explain the level of energy efficiency technology directly into the production frontier model. Originally, I planned to estimate the technology trends and then use the theoretical model to explain the changes in the trend. There is a bit of a disconnect in this paper between the theory which assumes that capital and energy are non-substitutable but the empirical model finds substantial substitutability. By that point it was too late and go back and develop a new theoretical model. This paper is way too big to submit to a journal (something like 19,000 words). So much of the literature review and theory will be cut or condensed in a journal submission. I have a bunch of stuff to do before that so it will be a while. I'll be happy to get feedback in the meantime. Especially, about any mistakes in the theoretical model.

Poverty and Progress: An Ecological Model of Economic Development

I remembered Mick Common mentioning this book and I saw that it was referred to by Robert Allen. I found that ANU's library had a copy but it was nowhere on the shelves and I put in a missing book search request. It couldn't be found. So I was surprised to get an e-mail while I was in Adelaide telling me to come pick it up from the library! You can see from the cover above that this book was published in 1973. Disappointingly, the copy I'm reading has a boring grey hardback cover.

The central idea of the book is similar to that of Ester Böserup - innovation and economic development are responses to the scarcity caused by increased population. This idea explains a lot, but I think that Wilkinson takes it too far. In his opinion economic development never increases welfare in the long-run. In his view all the innovations of modern industrial society are merely (often inferior) substitutes for goods that were lost in the industrialization process that was necessary to cope with increased population. If this were really true then why don't countries with low population densities relative to resources in today's world (New Zealand?) adopt a medieval way of life?

I think the idea does explain a lot about pre-industrial societies and the beginning of the industrial revolution. As Allen also documents, coal was an inferior fuel to wood, at least until innovations for using coal effectively came into play. Both authors agree that the scarcity of wood in England drove the increased adoption of coal for many uses.

Another important idea in the book is that many pre-industrial societies had various institutions to control population including taboos on sex at certain times or between certain classes of people, contraception methods, abortion, and infanticide. Some pre-industrial societies, therefore, managed to stay well within ecological bounds. Without facing the pressure of meeting subsistence needs there was little reason to innovate. When Christian missionaries arrived in many such places they tried to eliminate these institutions with a resultant take off of population growth.

The main methods in Christian Europe were delayed marriage in periods of reduced prosperity due to high population and (not mentioned by Wilkinson) monastic and priestly orders. These were less effective at maintaining the population within the carrying capacity for a comfortable lifestyle. England already reached carrying capacity in the 14th century. The Black Death then wiped out a large proportion of the population. Only in the 17th century was carrying capacity again approached. The eventual response was the agricultural and industrial revolutions. Would industrialization have started earlier if the Black Death plague hadn't happened?

Convergence of Energy Efficiency

I've been making more progress in my Environmental Economics Research Hub Project. I still don't have final results, but I am close. The above chart shows the estimated underlying energy efficiency, controlling for the structure of the economy etc. for five countries. I don't quite believe that the US is more efficient than Switzerland but at least the two developing economies are less efficient than the developed ones. However, it appears there has been convergence over time, particularly between China and the others. It's, therefore, not reasonable to suppose that gains in energy efficiency in China will be as fast in the future as in the 1979-2007 period under business as usual. Perhaps the 2000-2007 period will be more representative? This makes China's carbon intensity target a lot harder to achieve than many people think. It's not unachievable though given China's goals for renewable energy. This is what we are going to look at in our AARES paper.

Energy/Capital Ratio and Capital Density

Back in October, I blogged about the energy/capital ratio arguing that it was a rough proxy of the level of energy efficiency and/or environmental technology in a country. The following chart plots the energy/capital ratio against the capital density - the amount of capital per unit land area in a country (both are averages for 1971-2007):



Now there could be quite a lot of explanations of the strong correlation (-0.72) between the two variables including lower transport distances in more densely populated countries and measurement errors. But I'd argue that capital density is a rough proxy of the potential environmental disruption in the absence of any ameliorating policies and that E/K is a rough measure of the stringency of policy. In my 2005 paper in Journal of Environment and Development I argued that one of the key determinants of environmental policy was likely to be the level of damage in the absence of action, finding found some support for the hypothesis. I am now testing this idea more rigourously in my current research. So far it is holding up.

The important point is that K/T isn't just a function of the level of income per capita. Australia ($329k per km^2) and Canada ($362k per km^2) have similar levels of this variable. But so do Brazil and Peru. The USA has near $4 million per square kilometre, while Britain has $20 million...

Corruption and Development

Most of my readers are probably familiar with the correlation between corruption and the level of economic development, but as I was entering the data into my database I thought I'd go ahead and blog about it. The chart shows the 2007 Transparency International Corruption Perception Index on the Y axis and average income per capita over the 1971-2007 period in 2007 PPP dollars (from the Penn World Table) on the X axis for the 85 countries in my energy efficiency study. The sample excludes oil economies. The correlation between these variables in the sample is 0.90. It goes to 0.92 when Luxembourg - the point at the extreme right - is excluded. Including the petroleum exporters would reduce the correlation. The corruption index is bounded from above at 10.

The direction of causation is not clearcut. Poor countries tend to underpay their government employees (relative to the private sector) resulting in these employees seeking (small) bribes to make ends meet. In the other direction, corruption (large bribes) is supposed to result in economic policies that protect special interests and according to the Prescott-Parente theory of growth reduce total factor productivity and, therefore, GDP per capita in corrupt countries. My perception is that special interests in the energy and mining industries do wield a lot of influence in nominally low corruption economies such as the US and Australia. My colleague, Sambit Bhattacharya's research shows that resource endowments do not negatively affect growth in democracies. But do they negatively affect energy efficiency and environmental quality? In my Hub research, I am going to test the effects of resource endowments, perceived corruption and other variables on energy efficiency and hopefully come up with some answers.

Fenner Presentation Slides

Here are the slides for my presentation at the Fenner School today. Because of the way I set things up to allow me to have an "animation" in a pdf file there aren't actually as many slides as there are pages in this document. But I didn't delete the extra pages because this is my emergency copy if my flash drive fails!

My Fenner School Seminar: Thursday 12th November

As I mentioned a couple of months ago I'm giving a seminar at the Fenner School of the Environment and Society at ANU tomorrow, Thursday, at 1:00pm in the Forestry Lecture Theatre. As usual the slides will go up on the web as soon as I've completed them... I'm still working on model runs to put in the presentation as I wasn't pleased with the results I presented in Darwin. I'll show those results tomorrow and then some of the new ones. I'm also thinking about radically restructuring my Environmental Economics Research Hub project in order to be able to get it done in the time available. I would relegate the structural modeling I discussed earlier on this blog to future research and take a more reduced form approach here. There is plenty still to do on this project including the new obligation of producing a policy brief, which we'll present at AARES.

ANZSEE Presentation Slides

Here are the slides for my presentation at ANZSEE in Darwin on Wednesday. I only have 20 minutes to talk including questions so I'm going to have to cut something here. But thought I might as well put everything up for now. I'll be giving a longer version at the Fenner School at ANU on 12th November. Also coming up this month is a presentation at the ANU Economics Showcase 2009. I will be talking about the effect of sample size and estimator selection on what we think we know about economics - i.e how useful meta-analysis can be.

P.S. 11:45pm
I just cut a bunch of stuff out of my presentation for ANZSEE including all the early stuff about the EKC and put the shorter version up on the web.

A Statement of Your Most Significant Contributions to this Research Field

I have to write this for the grant applications I'm planning on filing early next year. Usually, in job applications you have to write about your current and planned research. But this is different and a bit weird. It feels a bit like I'm writing my own obituary :)

For U.S. grants you just send them your (condensed) CV. Australia is much more into giving money to individuals than for specific topics. I once had to write a "scientific autobiography" for a job I applied for in Israel. That was the same idea more or less.