Saturday, March 25, 2017

Adam Smith's ontology

In my last post, I had a list of Tony Lawson quotes critiquing modern mathematical economics. I believe if Adam Smith were alive today, he would support Lawson's message that economics needs to be based on ontology (the study of the nature and reality). Here is a quote from Lawson,
I'd like to state a case for ontology: the study of being. The study of nature and reality... Modern economics is unrealistic because most of the time economists are trying to make theories about a world that's open, structured, internally related and processual conform to a world that's atomistic and closed.
Written in 1776, The Wealth of Nations (Smith's magnum opus) is over 900 pages long. The Wealth of Nations explains economic phenomena with descriptive language and analyzes the causes and mechanism behind many different types of economic phenomena. Ultimately, The Wealth of Nations is about the nature of an economy. I believe this fits Lawson's description of what economists should be doing. Economic journalist David Warsh says,
[The Wealth of Nations] reads as much like a present-day business magazine as an eighteenth-century treatise on political economy... The book contains not a single chart, few enough numbers and no diagrams, but so penetrating was its reasoning that it launched a science. (The Knowledge and Wealth of Nations, 2006)
Given these two quotes, I believe Smith would support Lawson's philosophy. Lawson emphasizes non-mathematical approaches and abductive reasoning, which I believe The Wealth of Nations epitomizes.

Thursday, March 23, 2017

Tony Lawson's critique of mathematical economics

Tony Lawson is an economist best known for his critique on mathematical economics. This post is a collection of Tony Lawson quotes obtained from three lectures posted on YouTube. I recommend listening to his argument regarding what's wrong with economics.

"Many people think mathematics is essential for economics to be a science. I don't think it's necessary."

"I don't want to oppose mathematics in economics. It's the emphasis that mathematics is the proper way to do economics is what I'm critical of."

"I'd like to state a case for ontology: the study of being. The study of nature and reality."

"Mathematics is a tool just like any other. The mathematical deductive systems that economists use are tools and they have conditions under which they are appropriate."

"With ontology, we can get an insight to what sort of methods might be useful for [economics]."

"The essence of science is this move from phenomenon of interest to underlying cause. We can do this in the social realm because it's structured. We can go practices of interest to underlying cause. Practices bound up with poverty, crisis, whatever to underlying causes. This move is as open to us in economics as it is in any science."

"Deduction: all ravens are black, therefore the next raven I see is black. Induction: I see six black ravens, therefore all ravens are black. Retroduction is movement from the blackness to the cause of blackness."

"Modern economics is unrealistic because most of the time economists are trying to make theories about a world that's open, structured, internally related and processual conform to a world that's atomistic and closed."

Confronting Mathematical Models in Economics (posted June 2014)

"In economics faculties, probably more than 90% of what is taught focuses or employs some form of mathematical modeling."

"For most mathematical tools, a precondition for them being useful is that there are correlations. Correlations to be uncovered. Regularities for whenever X then Y."

"Closed systems are those in which correlations and event regularities occur and the ontology that is sufficient for it is a world of isolated atoms and that seems to be the implicit ontology of economics for the last 50 years."

"Mathematics is very good if you want to do deduction. Go from general to particular. If I say all ravens are black, I can deduce that the next raven I see is black. Mathematics also good for induction. If I've seen 100 black ravens and I infer all ravens are black, I can do it with mathematics."

"Underneath and grounding our social practices are social structures. For example social rules. This isn't new, but the claim I want to make is these rules are out of phase with our practices."

"In the social realm, all of its components are internally related to everything else."

"[The social realm] is reproduced and transformed through practice."

"Social reality is made up of communities where the components are often human beings and the rights, obligations and positions is what holds it together."

"When do we get systems of isolated systems in the social realm? Well, people get pretty atomistic when they're in traffic jams. That kind of behavior can probably be modeled mathematically."

What's Wrong with Modern Economics? (posted November 2015)

"A telescope, a microscope, an electron accelerator: they're all tools of analysis. Mathematical models are tools of analysis."

"The use of mathematics to do social analysis is a bit like using a hammer to cut the grass."

"I think pluralism is the right message. Pluralism isn't saying you're wrong. Pluralism is saying all voices ought to be listened to."

"Pluralism can't mean, or isn't very useful if it means anything goes. I take it pluralism means open to others... but it doesn't mean every theory is as good as every other theory."

"Most people responded to the crisis in economics by putting forward new theories, new policies, new models. What is missing is philosophical discussion."

"I'm going to suggest to you that [philosophical discussion] is fundamental. In fact all the problems of the discipline are basically ontological. The orientation of the discipline ought to be ontological as it is in physics, chemistry, biology and almost everything else."

"When you look to the physical realm, they don't really start with methods. They start with ontology. If they want to look at a planet far away, they build a telescope. If they want to check a theory about the nature of mass and they need to check out Higgs boson, they build an electron accelerator."

"Most problem solving occurs on the job. We need to know the nature of the problem we're dealing with."

"If it's the case you see forms of human behavior that seem to conform, then by all accounts model it mathematically."

"The goal is to see the structures that condition human interaction and identify those we don't want... Identify the structures that stand in the way to human flourishing."

Wednesday, March 15, 2017

Publishing and Promotion in Economics: the Curse of the Top 5

Earlier this year, there was a panel discussion between 5 influential economists talking about economic journals, citations and the incentives of economists. I highly recommend watching this conversation. The conversation included: James Heckman, George Akerlof, Angus Deaton, Drew Fudenberg and Lars Peter Hansen. This post is a collection quotes from the conversation.

James Heckman (2000 Nobel Prize recipient)

"When I go to dinner with a group of young pre-tenure PhD economists or talk to graduate students who are thinking about doing research, I find increasingly, and even with post-tenure economists, there seems to be an obsession on ranking themselves by the number of top 5 journals articles published."

"If people are really aiming to go just into the top 5 journals and their goal is to publish papers as opposed to contribute to basic knowledge, I think that's a limited vision of economics and what economic research should be."

George Akerlof (2001 Nobel Prize recipient)

"Economists know there's a basic problem with any fixed reward system: people work to the test."

"People divide their effort between tasks according to the respective rewards of those tasks. If the rewards are to publish in the top 5, people will put their effort into whatever it takes to publish there. But that may not be the effort that makes them the best economist. Furthermore, in real life the game does not stop there. There's a doubly bad equilibrium because those who are successful at the task are those who then set the reward system in the future."

"If there is low power of tests for what is good or bad science, there are equilibrium traps. In these traps, new and better paradigms will fail to be adopted. On the contrary, such bad equilibrium traps disappear if there are high power of tests."

"There's almost a total disconnect between sociology and economics. But are they always wrong and we always right? Isn't it some combination of the two more likely to be correct and also relevant for the types of situations we as economists tend to look at?"

"There's a whole type of economics that economists simply don't do that doesn't occur in our journals. For examples economists did not provide detailed reportage of what was happening in the financial system [in 2005]."

"The authors should own the paper. When we submit the paper it should be our paper. We shouldn't have to rewrite it for whatever the referee thinks the paper should be."

"What I'm worried most of all, is what we don't see. I'm worried about the analysis that is never seen, that never becomes a paper and it doesn't become a paper because it can't be a paper. It can't become a paper because that's not what a paper in economics is all about."

"You usually don't know such vacuums exist. But there's actually a signal that such a black hole does exist because there wasn't anything in the economics profession that was detailing what was going to happen when we got into the great recession."

"I believe we should allow stuff that is much fuzzier so people can say what's in their hearts and that way may fill in some of these vacuums that are now black holes."

Angus Deaton (2015 Nobel Prize recipient)

"The top 5 journals are not really much about communicating new findings, that's a shame in some ways."

"There's been a large increase in evaluation by metrics, whether within departments or across departments by government agencies. In those, international journals are given very high weight."

"When somebody tells me, 'We'll that's your favorite candidate for bringing here, but we don't like their work at all' and you can try to explain but often you get nowhere at all because there isn't a common ground for communicating that information... That's why these algorithms, and this very silly algorithm (top 5) has become very popular."

"I think more sophisticated algorithms, if departments used them, deans made departments use them, would actually weaken the position of top journals. So what we have now, counting top journal citations is really stupid I think."

"I think that thinking about algorithms would be really helpful, even if they are very much a second best solution."

Drew Fudenberg

"We're not alone in this, we didn't just for some bizarre reason switch to a bad equilibrium. For some reason parallel things are happening in other fields."

"Since I joined this field, people have complained that papers are too long and are getting longer."

"Another trend in biology, which I think is going on in economics, is more co-authors per paper."

"I'd like to see journals have more papers in their lead journal but also have more subsidiary journals. I believe that's one way forward."

Lars Peter Hansen (2013 Nobel Prize recipient)

"This reliance on referees leads to a much more conservative strategy. I think it works against novel papers that cross subfield boundaries and that makes it all the more challenging. Basically it makes the simplest path to publication in the top 5 journals to be high quality follow up papers."

"I think it's hard to use citation counts in sensible ways that are applicable for young scholars."

"Sometimes there are papers that basically close a field they are so exceptional. They basically clear up everything so they don't have a whole lot of follow up work and therefore lose the citations."

"I don't see simple algorithmic alternatives as a way to evaluate young people, but I do think figuring out ways to encourage synergistic activities across subfields and to figure out ways that recognize creativity and possibilities of it at early stages of a career is absolutely vital and important."

Monday, March 13, 2017

Macroeconomic diagrams

This post is a collection of diagrams intended to display the macroeconomy. Diagrams have been used since ancient times to describe complex phenomena. In 1996, Bert Hall said, "diagrams are simplified figures, caricatures in a way, intended to convey essential meaning." I included my own version at the end of the post.

Richard Cantillon (1680's-1734)

Francois Quesnay (1694-1774)

Paul Samuelson (1915-2009)

Ari89 (Wikipedia username)

Irconomics (Wikipedia username)

Econ Analysis Tools

Saturday, March 11, 2017

Russ Roberts and empricial economics

There is currently an interesting conversation going on regarding empirical economics and predicting the future. This conversation was sparked from Russ Robert's recent essay titled 'What do Economists know'. In this essay, Roberts is skeptical of econometric analysis.

In the blog post Roberts says,
One response to the questions I am raising is that we have new techniques that solve a lot of the problems I’m talking about. These new empirical techniques have allowed us to run quasi-experiments that while not perfect, eliminate many of the problems of complexity and multi-causal reality. I remain a skeptic. But maybe the champions of the new empirical techniques will convince the skeptics. We’ll see... We are notoriously unreliable at the things the world really cares about and asks of our profession... What I am arguing here is that the combination of economics with statistics in a complex world promises a lot more than it delivers. We economists should be more humble and honest about the reliability and precision of statistical analysis.
John Cochrane praises Roberts' blog post. Cochrane says,
Russ explains with great clarity, just how uncertain [economic] estimates are... Economic history teaches us humility: No economist in 1900 could have figured out what farmers, horse-shoers, ice deliverers, street-sweepers, and so forth would do when those jobs disappeared. The people involved did. Knowledge of our own ignorance is useful.
There has also been some pushback to the post. Adam Ozimek says,
I think Russ is most correct in arguing for more humility from economists. But I would add that this is a general problem not particular to economists, and I would conjecture that on average those who do not attempt to tie down their beliefs with empirical research suffer from an even greater humility gap than those who do. Within economics, failure to regularly cite empirical results, or lack of familiarity with them, seems pretty strongly correlated with lack of humility and not vice versa.
Noah Smith also provides a critique on Roberts' post,
The alternative to empiricism in economics is not agnostic humility, but intuitionism - the idea that we can know about the world by thinking about how it works, and that exposure to evidence will only pollute the truths that we divine from our own minds. And that's something I think economists need to avoid.
I agree the most with Roberts and Cochrane. I believe one of the main goals of macroeconomics is to predict future economic phenomena. Inevitably, predictions far into the future (such as 10 or 30 years) require knowledge about new technologies and future political institutions. This is why macroeconomics is so difficult. I doubt modern econometric analysis can predict future technologies and future political institutions.

I'm not saying that macroeconomics should abandon the goal of becoming an objective science. I believe that there are macroeconomic consistencies that can be supported with evidence. But predicting what the actual economy will look like in 20 years is outside the realm of modern econometrics. Predicting things like this relies on a different set of knowledge and ability.

Friday, March 10, 2017

Predictive power and macroeconomics

I believe macroeconomics has three main components: past, present and future. Macroeconomists should have detailed knowledge about economic history, be able to explain economic phenomena and be able to predict the future economy. An unusually gifted macroeconomist should be able to predict many different macroeconomic variables (employment, industries, companies, trade, etc.) for many different countries over short and long time horizons. Nobel laureate Paul Samuelson once said in an interview,
My notion of a fruitful economic science would be that it can help us explain and understand the course of actual economic history. A scholar who seriously addresses commentary on contemporary monthly and yearly events is, in this view, practicing the study of history—history in its most contemporary time phasing.
I believe there should be a standard test for macroeconomists to predict a collection of about 40 statistics for the economy (for example, GDP growth, inflation and exports). The weighted sum of errors for all the predictions would be complied to determine a numerical grade for the macroeconomist. In other words, macroeconomists should try to predict the future statistics on Econ Factbook. This test would be similar to an IQ test for macroeconomics.

What tools would a macroeconomist use to for this test? I think the best predictions in the short term would be based on mathematical models (such as DSGE) but predictions further into the future would rely more on judgment calls. Predictions far into the future (such as 10 or 30 years) require knowledge about new technologies, future political institutions, future consumer demand and the evolution of society. This is why macroeconomics is so difficult.

I think this test would reveal which theoretical systems are most effective for macroeconomics. Maybe economists who specialize in economic history would perform the best. Maybe the econometricians would perform the best. I'm not sure who would produce the most accurate predictions, but I think this would be a productive exercise for macroeconomists.